The prevalence of type 2 diabetes mellitus (T2DM) is increasing exponentially worldwide. According to the Centers for Disease Control and Prevention, more than 23 million Americans had diabetes in 2007.¹ Globally, the prevalence of diabetes, of which T2DM accounts for 90% to 95% of cases,¹ is expected to increase from 171 million in 2000 to 366 million in 2030.² The National Health and Nutrition Examination Survey (NHANES) showed that about 66% of Americans were overweight or obese between 2003–2004.³ Data from a Swedish National Diabetes Register study showed both overweight and obesity as independent risk factors for cardiovascular disease (CVD) in patients with T2DM.⁴

This article presents an overview of the evolving concepts of the pathophysiology of T2DM, with a focus on two new therapeutic classes: the glucagon-like peptide–1 (GLP-1) receptor agonists and the dipeptidyl peptidase–4 (DPP-4) inhibitors.

THE PATHOPHYSIOLOGY OF T2DM

The American Association of Clinical Endocrinologists (AACE) describes T2DM as “a progressive, complex metabolic disorder characterized by coexisting defects of multiple organ sites including insulin resistance in muscle and adipose tissue, a progressive decline in pancreatic insulin secretion, unrestrained hepatic glucose production, and other hormonal deficiencies.”⁵ Other defects include accelerated gastric emptying in patients with T2DM, especially those who are obese or who have the disease for a long duration.^6,7

Hormonal deficiencies in T2DM are related to abnormalities in the secretion of the beta-cell hormone amylin, the alpha-cell hormone glucagon, and the incretin hormones GLP-1 and glucose-dependent insulinotropic polypeptide (GIP).^8,9 In addition to the triumvirate of core defects associated with T2DM (involvement of the pancreatic beta cell, muscle, and liver), other mechanisms of disease onset have been advanced, including accelerated lipolysis, hyperglucagonemia, and incretin deficiency/resistance.⁹ Also, the rate of basal hepatic glucose production is markedly increased in patients with T2DM, which is closely correlated with elevations in fasting plasma glucagon concentration.⁹

The incretin effect—the intestinal augmentation of secretion of insulin—attributed to GLP-1 and GIP is reduced in patients with T2DM.¹⁰ The secretion of GIP may be normal or elevated in patients with T2DM while the secretion of GLP-1 is deficient; however, cellular responsiveness to GLP-1 is preserved while responsiveness to GIP is diminished.¹¹

Both endogenous and exogenous GLP-1 and GIP are degraded in vivo and in vitro by the enzyme DPP-4,¹²
a ubiquitous, membrane-spanning, cell-surface aminopeptidase that preferentially cleaves peptides with a proline or alanine residue in the second amino-terminal position. DPP-4 is widely expressed (eg, in the liver, lungs, kidney, lymphocytes, epithelial cells, endothelial cells). The role of DPP-4 in the immune system stems from its exopeptidase activity and its interactions with various molecules, including cyto­kines and chemokines.¹³

INCRETIN-BASED THERAPIES: GLP-1 RECEPTOR AGONISTS AND DPP-4 INHIBITORS

Exenatide is a GLP-1 receptor agonist that is resistant to DPP-4 degradation. Based on preclinical studies, exenatide, which shares a 53% amino acid sequence identity with human GLP-1, is approximately 5,500 times more potent than endogenous GLP-1 in glucose lowering.^14,15 Among the acute actions of exenatide is glucose-dependent insulinotropism, the end result of which may be a reduced risk of hypoglycemia.¹⁶ This contrasts with insulin secretagogues (eg, sulfonylureas), which increase insulin secretion regardless of glucose concentrations.

Exenatide received US Food and Drug Administration (FDA) approval in 2005 and is indicated for the treatment of patients with T2DM.^13,17 Exenatide is administered BID as a subcutaneous (SC) injection in doses of 5 or 10 μg within 1 hour before the two major meals of the day, which should be eaten about 6 hours apart.¹⁸

Approved in 2006, sitagliptin was the first DPP-4 inhibitor indicated for adjunctive therapy to lifestyle modifications for the treatment of patients with T2DM.¹⁷ The recommended dosage of oral sitagliptin is 100 mg QD. A single-tablet formulation of the combination of sitagliptin and metformin was approved by the FDA in 2007.¹⁹ Another DPP-4 inhibitor, saxagliptin, was approved in July 2009 for treatment of patients with T2DM either as monotherapy or in combination with metformin, sulfonylurea, or a thiazolidinedione (TZD).²⁰ The DPP-4 inhibitor vildagliptin is approved in the European Union and Latin America but not in the United States. Vildagliptin is available as a 50- or 100-mg daily dosage; it has been recommended for use at 50 mg QD in combination with a sulfonylurea or at 50 mg BID with either metformin or a TZD.¹⁸

GLP-1 RECEPTOR AGONISTS AND DPP-4 INHIBITORS IN DEVELOPMENT

Exenatide is currently being evaluated as a once-weekly formulation.^21,22 Compared with the BID formulation, exenatide once weekly has been shown to produce significantly greater improvements in glycemic control, with similar reductions in body weight and no increased risk of hypoglycemia.²¹

Also undergoing regulatory review is the partly DPP-4–resistant acylated GLP-1 receptor agonist liraglutide.¹³ Liraglutide, a human analogue GLP-1 receptor agonist, has 97% linear amino acid sequence homology to human GLP-1.^23,24 Based on its prolonged degradation time and resulting 10- to 14-hour half-life, liraglutide is anticipated to be dosed once daily.^13,25,26

Other GLP-1 receptor agonists and DPP-4 inhibitors are in varying stages of development.²⁷ Albiglutide is a long-acting GLP-1 receptor agonist that is generated by the genetic fusion of a DPP-4–resistant GLP-1 to human albumin. Based on pharmacokinetic studies, albiglutide has a half-life of 6 to 8 days. AVE0010, an exendin-4-based GLP-1 receptor agonist, was shown in a 28-day T2DM clinical trial to have an affinity four times greater than native GLP-1 for the human GLP-1 receptor.²⁷ Taspoglutide (R1583), a human analogue GLP-1 receptor agonist, was evaluated in three randomized, placebo-controlled studies as a GLP-1 receptor agonist. Alogliptin, a DPP-4 inhibitor currently in development, has been shown to be safe and effective in studies as monotherapy and in combination with other antidiabetes agents.^28–30

 

 

CLINICAL TRIALS: GLP-1 RECEPTOR AGONISTS AND DPP-4 INHIBITORS

This section summarizes clinical trials of GLP-1 receptor agonists and DPP-4 inhibitors. The summary is based on literature published from 2005 to 2009 relevant to phase 3 or 4 T2DM clinical trials with currently available agents, or agents with pending new drug applications.

Table 1 summarizes the data on the effects of the GLP-1 receptor agonists on glucose lowering based on glycosylated hemoglobin (HbA1c) mean changes from baseline, body weight, and hypoglycemia. Eleven studies were identified for exenatide, including three pivotal trials,^31–33 three insulin-comparator studies,^34–36 one long-term study,³⁷ one monotherapy study (a use for which it is not currently indicated),³⁸ one head-to-head study with a DPP-4 inhibitor,³⁹ and two studies with exenatide once weekly (which is currently investigational).^21,22 Five primary efficacy studies with liraglutide were also identified.^{23,25,26,40,41}

Table 2 summarizes the corresponding data for the DPP-4 inhibitors. Ten studies with sitagliptin were identified, including four monotherapy studies,^42–45 one head-to-head study with a GLP-1 receptor agonist,³⁹ and five studies in which sitagliptin was used in combination or as add-on therapy.^46–50 Five saxagliptin studies are reviewed, including two in which saxagliptin was used in combination with metformin and one in combination with glyburide.^51–55 Six studies with vildagliptin were reviewed,^56–61 but no trials specific to the single-tablet formulation of sitagliptin plus metformin were identified.

Effects on HbA1c and weight

GLP-1 receptor agonists reduced HbA1c. Based on the studies reviewed in Table 1, exenatide BID reduced baseline HbA1c by a maximum of –1.5% at 30 weeks.^21,31,32 Exenatide has demonstrated sustained reductions in HbA1c of –0.8% for up to 3.5 years in an open-label extension trial.³⁷ Even greater reductions in HbA1c (–1.4% at 15 weeks and –1.9% at 30 weeks) have been reported with the once-weekly formulation under clinical development.^21,22 Liraglutide, another GLP-1 receptor agonist under development, has reported HbA1c reductions from baseline up to –1.67% at 14 weeks,^40,41 up to –1.1% at 26 weeks,^23,26 and up to –1.14% at 52 weeks.²⁵ The reductions quoted generally refer to means, and individual patients may have greater or lesser responses. Also, baseline HbA1c is a significant determinant of the potential HbA1c reduction. Higher baseline values drop more significantly than do baseline values that are closer to normal.

Weight reduction with GLP-1 receptor agonists. In addition to effective glucose lowering, the GLP-1 receptor agonists, particularly exendin-4 agonists, produced beneficial effects on weight (Table 1). Exenatide BID elicited mean weight reductions up to –3.6 kg at 30 weeks^21,31,32 and –5.3 kg at 3.5 years.³⁷ Exenatide once weekly resulted in mean weight reductions of up to –3.8 kg at 15 weeks²² and –3.7 kg at 30 weeks.²¹ Effects on weight with liraglutide varied from a mean reduction of up to –2.99 kg to a slight gain of up to +0.13 kg at 14 weeks^40,41 and with weight loss of up to –2.8 kg at 26 weeks^23,26 and up to –2.5 kg at 52 weeks.²⁵ In this review, only exenatide has been assessed in insulin-comparator studies, where it was shown to reduce weight compared with the insulin analogues, which led to weight gain.^34–36

Hypoglycemia. Patients receiving exenatide experienced lower rates of hypoglycemia (up to 17%) than patients treated with either insulin glargine or insulin aspart (~25%).^34,36 The rate of hypoglycemia with exenatide is comparable to that seen with metformin (up to 21%) in a systematic review of oral antidiabetes agents conducted by the Agency for Healthcare Research and Quality.⁶² No major hypoglycemic events were reported in the liraglutide studies reviewed. The incidence of hypoglycemia reported with DPP-4 inhibitors (Table 2) is also low (2% or less in most studies). The glucose-dependent mechanisms of the incretin-based therapies minimizes the risk of hypoglycemia.

DPP-4 inhibitors and sustained HbA1c reduction. The effects of the DPP-4 inhibitors on HbA1c and weight, either as monotherapy or in combination with other agents, were evaluated in studies ranging in duration from 12 to 52 weeks (Table 2). No studies were identified that compared the glycemic control effects of DPP-4 inhibitors and insulin analogues. Sitagliptin led to a mean reduction in HbA1c from baseline of up to –0.65% at 12 weeks,^43,45 up to –0.48% at 18 weeks,⁴⁴ up to –0.85% at 24 weeks,^42,46,47,50 up to –1.0% at 30 weeks,⁴⁹ and up to –0.67% at 52 weeks.⁴⁸ Saxagliptin mean reductions in HbA1c ranged from –0.43% to –1.17%.^51–54 Data from four 24-week T2DM studies^56–60 showed vildagliptin reducing HbA1c up to –1.4% at 24 weeks, with the greatest reduction in a study that involved drug-naïve patients with a relatively short duration of disease (mean, 1.2 years).⁵⁹ Reductions in HbA1c of –1.0% were sustained in a 52-week study⁶¹ and its 52-week extension.⁵⁸

DPP-4 inhibitors: weight neutral. The DPP-4 inhibitors appear to have a weight-neutral effect (Table 2). The effects of sitagliptin on weight ranged from a loss of –1.5 kg⁴⁸ at 52 weeks to a gain of +1.8 kg at 24 weeks.⁵⁰ Weight changes with saxagliptin ranged from a mean reduction of –1.8 kg⁵³ to a gain of +0.7 kg.⁵¹ Two vildagliptin studies showed varying effects on weight ranging from a loss of up to –1.8 kg from baseline⁵⁶ to a gain of up to +1.3 kg⁵⁷ relative to placebo, both at 24 weeks.

Potential for CV risk reduction

Potentially beneficial effects on CV risk factors, including blood pressure (ie, reduction) and lipid concentrations (ie, differential effects on low-density lipoprotein and high-density lipoprotein cholesterol), were identified in seven GLP-1 receptor studies—three with exenatide (two with exenatide BID,^37,38 and one with the investigational exenatide once weekly²¹) and four with liraglutide.^23,25,26,41 For the DPP-4 inhibitors, three studies were identified—two with sitagliptin^45,50 and one with vildagliptin⁶¹—in which potentially beneficial effects on CV risk factors were demonstrated.The data have been encouraging, although the clinical implications have yet to be fully understood.

Head-to-head comparison

A recent study compared the effects of the GLP-1 receptor agonist exenatide and the DPP-4 inhibitor sitagliptin on postprandial glucose (PPG) concentrations, insulin and glucagon secretion, gastric intake, and caloric intake.³⁹ Although limited by a short treatment duration (2 weeks), the study showed that the GLP-1 receptor agonist had a greater effect than the DPP-4 inhibitor in reducing PPG concentrations, a more potent effect in increasing insulin secretion and decreasing postprandial glucagon secretion, and a relatively greater effect in reducing caloric intake; and that it decreased the rate of gastric emptying (sitagliptin had no effect). These differences suggest that exenatide may provide a greater degree of GLP-1 receptor activation than the more physiologic concentrations of GLP-1 reached with DPP-4 inhibition.³⁹ Results of a scintigraphic study showed that exenatide substantially slows the gastric emptying that is accelerated in patients with T2DM. This could be another beneficial mechanism in treating postprandial glycemia.⁶³

Adverse effects

Exenatide has shown effects on hepatic injury markers (ie, improvement in alanine and aspartate aminotransferases) for up to 3.5 years of treatment.³⁷ For the GLP-1 receptor agonist and DPP-4 inhibitor studies reviewed, the adverse events were generally mild and included nausea and vomiting, nasopharyngitis, and mild hypoglycemia.

 

 

Meta-analysis conclusions

The published clinical trial data presented in this review expand the body of evidence on the safety and efficacy of incretin-based therapy in patients with T2DM. These data include the results of a meta-analysis by Amori et al,¹⁷ which examined randomized controlled trials of 12 weeks’ or longer duration that compared incretin-based therapy with placebo or other diabetes medications and reported HbA1c changes in adults with T2DM. The meta-analysis showed that incretin-based therapies reduced HbA1c more than placebo (weighted mean difference, –0.97% [95% confidence interval (CI), –1.13% to –0.81%] for GLP-1 receptor agonists and –0.74% [95% CI, –0.85% to –0.62%] for DPP-4 inhibitors) and were noninferior to other antidiabetes agents. Treatment with a GLP-1 receptor agonist (ie, exenatide) caused weight loss (–1.4 kg and –4.8 kg vs placebo and insulin, respectively) while DPP-4 inhibitors (ie, sitagliptin, vildagliptin) were weight neutral.¹⁷

Beta-cell function

Evidence regarding the effects of incretin-based therapies, particularly the exendin-4 GLP-1 receptor agonists, on beta-cell function in patients with T2DM continues to accumulate. When assessing long-term (1 year) exenatide treatment in patients with T2DM, a trial (n = 69) comparing exenatide with the basal insulin analogue insulin glargine showed that exenatide and insulin glargine resulted in similar reductions in HbA1c (–0.8% vs –0.7%; P = .55).⁶⁴ However, exenatide significantly reduced body weight while insulin glargine resulted in weight gain (–3.6 kg vs +1.0 kg; P < .0001). In terms of beta-cell function, arginine-stimulated C-peptide secretion during hyperglycemia increased 2.46-fold from baseline after 52 weeks of exenatide treatment compared with 1.31-fold with insulin glargine treatment (P < .0001).⁶⁴

With respect to the direct beta-cell effects of liraglutide, a preclinical study reported that liraglutide improved glucose homeostasis in marginal mass islet transplantation in diabetic mice.⁶⁵ In this study, liraglutide was shown, in a mouse model, to reduce the time to normoglycemia after islet cell transplantation (median time, 1 vs 72.5 days; P < .0001). The effects of liraglutide on beta-cell function also were assessed in 13 patients with T2DM. After 7 days of treatment, liraglutide improved beta-cell function, which was associated with improvement in glucose concentration.⁶⁶ Liraglutide improved potentiation of insulin secretion during the first meal, owing in part to restoration of the potentiation peak (which is markedly blunted in T2DM), in a phenomenon similar to that observed with exenatide.⁶⁷

Beneficial effects on beta-cell function have also been reported with DPP-4 inhibitors. In a model-based analysis of patients with T2DM, it was shown that sitagliptin improved basal, static, and dynamic responsiveness of pancreatic beta cells to glucose. The results were observed when sitagliptin was administered both as an add-on to metformin therapy and as monotherapy.⁶⁸ A 52-week, double-blind, randomized, parallel-group study compared vildagliptin 50 mg/day and placebo in 306 patients with T2DM and mild hyperglycemia (HbA1c, 6.2% to 7.5%). Vildagliptin was shown to significantly increase fasting insulin secretory tone, glucose sensitivity, and rate sensitivity, all of which are aspects of beta-cell function.⁶⁹

Summary

Based on the ability of incretin-based therapies to address various disease mechanisms, including beta-cell defects (ie, hyperglycemia), hormone-related abnormalities (ie, hyperglucagonemia, incretin deficiency/resistance), and accelerated gastric emptying (especially with GLP-1 receptor agonists); their favorable effects on weight (reduction with GLP-1 receptor agonists and neutral with DPP-4 inhibitors); their beneficial effects on CV risk factors; and their good safety profile (ie, hypoglycemia risk comparable with metformin), these agents could be considered therapeutic advances for the treatment of patients with T2DM.

INCRETIN-BASED THERAPIES IN GUIDELINES AND ALGORITHMS

The 2007 AACE medical guidelines for clinical practice for the management of diabetes recognized the place of the incretin-based therapies and included them among the pharmacologic options.⁵ Exenatide was specifically recommended for combination therapy with metformin, a sulfonylurea (secretagogue), a sulfonylurea plus metformin, or a TZD. Sitagliptin was recommended for use as monotherapy or in combination with metformin or a TZD.⁵

In 2009, the American Diabetes Association (ADA) and the European Association for the Study of Diabetes convened a consensus panel to produce an algorithm for the initiation and adjustment of therapy for patients with T2DM. In this algorithm, GLP-1 receptor agonists were considered appropriate in certain clinical scenarios (eg, when hypoglycemia was an issue or weight loss was a major consideration during treatment). However, the groups also noted a need for more data on long-term safety and the cost of treatment with incretin-based therapies.⁷⁰

The AACE and the American College of Endocrinology recently developed “road maps” for managing patients with T2DM. In patients with T2DM who are naïve to therapy, DPP-4 inhibitors are among the recommended first options when the initial HbA1c is 6.0% to 7.0% and as a combination therapy component when HbA1c reaches 7.0% to 9.0%. In patients who have already received monotherapy for 2 to 3 months and whose HbA1c is 6.5% to 8.5%, treatment options include combination therapy with a DPP-4 inhibitor and metformin or a TZD. Another option includes the initiation of treatment with a GLP-1 receptor agonist in combination with a TZD, with metformin or a sulfonylurea, or with metformin and a sulfonylurea.⁷¹

The role of GLP-1 receptor agonist therapies and their incorporation into T2DM treatment algorithms was noted at the 2008 annual meeting of the ADA. In the Banting lecture, Ralph A. DeFronzo, MD, advocated the early use of triple-drug therapy with metformin, exenatide, and a TZD in the management of patients with T2DM.⁹

CONCLUSION

T2DM, which is linked to weight gain and obesity, is a complex disease that predisposes patients to and is associated with CVD. A better understanding and appreciation of the role of the incretin system in the pathogenesis of T2DM has led to the development of incretin-based therapies, such as the GLP-1 receptor agonists and DPP-4 inhibitors. As more experimental and clinical evidence becomes available, subtle nuances are emerging that distinguish the roles of these two therapeutic classes.

The prevalence of type 2 diabetes mellitus (T2DM) is increasing exponentially worldwide. According to the Centers for Disease Control and Prevention, more than 23 million Americans had diabetes in 2007.¹ Globally, the prevalence of diabetes, of which T2DM accounts for 90% to 95% of cases,¹ is expected to increase from 171 million in 2000 to 366 million in 2030.² The National Health and Nutrition Examination Survey (NHANES) showed that about 66% of Americans were overweight or obese between 2003–2004.³ Data from a Swedish National Diabetes Register study showed both overweight and obesity as independent risk factors for cardiovascular disease (CVD) in patients with T2DM.⁴

This article presents an overview of the evolving concepts of the pathophysiology of T2DM, with a focus on two new therapeutic classes: the glucagon-like peptide–1 (GLP-1) receptor agonists and the dipeptidyl peptidase–4 (DPP-4) inhibitors.

THE PATHOPHYSIOLOGY OF T2DM

The American Association of Clinical Endocrinologists (AACE) describes T2DM as “a progressive, complex metabolic disorder characterized by coexisting defects of multiple organ sites including insulin resistance in muscle and adipose tissue, a progressive decline in pancreatic insulin secretion, unrestrained hepatic glucose production, and other hormonal deficiencies.”⁵ Other defects include accelerated gastric emptying in patients with T2DM, especially those who are obese or who have the disease for a long duration.^6,7

Hormonal deficiencies in T2DM are related to abnormalities in the secretion of the beta-cell hormone amylin, the alpha-cell hormone glucagon, and the incretin hormones GLP-1 and glucose-dependent insulinotropic polypeptide (GIP).^8,9 In addition to the triumvirate of core defects associated with T2DM (involvement of the pancreatic beta cell, muscle, and liver), other mechanisms of disease onset have been advanced, including accelerated lipolysis, hyperglucagonemia, and incretin deficiency/resistance.⁹ Also, the rate of basal hepatic glucose production is markedly increased in patients with T2DM, which is closely correlated with elevations in fasting plasma glucagon concentration.⁹

The incretin effect—the intestinal augmentation of secretion of insulin—attributed to GLP-1 and GIP is reduced in patients with T2DM.¹⁰ The secretion of GIP may be normal or elevated in patients with T2DM while the secretion of GLP-1 is deficient; however, cellular responsiveness to GLP-1 is preserved while responsiveness to GIP is diminished.¹¹

Both endogenous and exogenous GLP-1 and GIP are degraded in vivo and in vitro by the enzyme DPP-4,¹²
a ubiquitous, membrane-spanning, cell-surface aminopeptidase that preferentially cleaves peptides with a proline or alanine residue in the second amino-terminal position. DPP-4 is widely expressed (eg, in the liver, lungs, kidney, lymphocytes, epithelial cells, endothelial cells). The role of DPP-4 in the immune system stems from its exopeptidase activity and its interactions with various molecules, including cyto­kines and chemokines.¹³

INCRETIN-BASED THERAPIES: GLP-1 RECEPTOR AGONISTS AND DPP-4 INHIBITORS

Exenatide is a GLP-1 receptor agonist that is resistant to DPP-4 degradation. Based on preclinical studies, exenatide, which shares a 53% amino acid sequence identity with human GLP-1, is approximately 5,500 times more potent than endogenous GLP-1 in glucose lowering.^14,15 Among the acute actions of exenatide is glucose-dependent insulinotropism, the end result of which may be a reduced risk of hypoglycemia.¹⁶ This contrasts with insulin secretagogues (eg, sulfonylureas), which increase insulin secretion regardless of glucose concentrations.

Exenatide received US Food and Drug Administration (FDA) approval in 2005 and is indicated for the treatment of patients with T2DM.^13,17 Exenatide is administered BID as a subcutaneous (SC) injection in doses of 5 or 10 μg within 1 hour before the two major meals of the day, which should be eaten about 6 hours apart.¹⁸

Approved in 2006, sitagliptin was the first DPP-4 inhibitor indicated for adjunctive therapy to lifestyle modifications for the treatment of patients with T2DM.¹⁷ The recommended dosage of oral sitagliptin is 100 mg QD. A single-tablet formulation of the combination of sitagliptin and metformin was approved by the FDA in 2007.¹⁹ Another DPP-4 inhibitor, saxagliptin, was approved in July 2009 for treatment of patients with T2DM either as monotherapy or in combination with metformin, sulfonylurea, or a thiazolidinedione (TZD).²⁰ The DPP-4 inhibitor vildagliptin is approved in the European Union and Latin America but not in the United States. Vildagliptin is available as a 50- or 100-mg daily dosage; it has been recommended for use at 50 mg QD in combination with a sulfonylurea or at 50 mg BID with either metformin or a TZD.¹⁸

GLP-1 RECEPTOR AGONISTS AND DPP-4 INHIBITORS IN DEVELOPMENT

Exenatide is currently being evaluated as a once-weekly formulation.^21,22 Compared with the BID formulation, exenatide once weekly has been shown to produce significantly greater improvements in glycemic control, with similar reductions in body weight and no increased risk of hypoglycemia.²¹

Also undergoing regulatory review is the partly DPP-4–resistant acylated GLP-1 receptor agonist liraglutide.¹³ Liraglutide, a human analogue GLP-1 receptor agonist, has 97% linear amino acid sequence homology to human GLP-1.^23,24 Based on its prolonged degradation time and resulting 10- to 14-hour half-life, liraglutide is anticipated to be dosed once daily.^13,25,26

Other GLP-1 receptor agonists and DPP-4 inhibitors are in varying stages of development.²⁷ Albiglutide is a long-acting GLP-1 receptor agonist that is generated by the genetic fusion of a DPP-4–resistant GLP-1 to human albumin. Based on pharmacokinetic studies, albiglutide has a half-life of 6 to 8 days. AVE0010, an exendin-4-based GLP-1 receptor agonist, was shown in a 28-day T2DM clinical trial to have an affinity four times greater than native GLP-1 for the human GLP-1 receptor.²⁷ Taspoglutide (R1583), a human analogue GLP-1 receptor agonist, was evaluated in three randomized, placebo-controlled studies as a GLP-1 receptor agonist. Alogliptin, a DPP-4 inhibitor currently in development, has been shown to be safe and effective in studies as monotherapy and in combination with other antidiabetes agents.^28–30

 

 

CLINICAL TRIALS: GLP-1 RECEPTOR AGONISTS AND DPP-4 INHIBITORS

This section summarizes clinical trials of GLP-1 receptor agonists and DPP-4 inhibitors. The summary is based on literature published from 2005 to 2009 relevant to phase 3 or 4 T2DM clinical trials with currently available agents, or agents with pending new drug applications.

Table 1 summarizes the data on the effects of the GLP-1 receptor agonists on glucose lowering based on glycosylated hemoglobin (HbA1c) mean changes from baseline, body weight, and hypoglycemia. Eleven studies were identified for exenatide, including three pivotal trials,^31–33 three insulin-comparator studies,^34–36 one long-term study,³⁷ one monotherapy study (a use for which it is not currently indicated),³⁸ one head-to-head study with a DPP-4 inhibitor,³⁹ and two studies with exenatide once weekly (which is currently investigational).^21,22 Five primary efficacy studies with liraglutide were also identified.^{23,25,26,40,41}

Table 2 summarizes the corresponding data for the DPP-4 inhibitors. Ten studies with sitagliptin were identified, including four monotherapy studies,^42–45 one head-to-head study with a GLP-1 receptor agonist,³⁹ and five studies in which sitagliptin was used in combination or as add-on therapy.^46–50 Five saxagliptin studies are reviewed, including two in which saxagliptin was used in combination with metformin and one in combination with glyburide.^51–55 Six studies with vildagliptin were reviewed,^56–61 but no trials specific to the single-tablet formulation of sitagliptin plus metformin were identified.

Effects on HbA1c and weight

GLP-1 receptor agonists reduced HbA1c. Based on the studies reviewed in Table 1, exenatide BID reduced baseline HbA1c by a maximum of –1.5% at 30 weeks.^21,31,32 Exenatide has demonstrated sustained reductions in HbA1c of –0.8% for up to 3.5 years in an open-label extension trial.³⁷ Even greater reductions in HbA1c (–1.4% at 15 weeks and –1.9% at 30 weeks) have been reported with the once-weekly formulation under clinical development.^21,22 Liraglutide, another GLP-1 receptor agonist under development, has reported HbA1c reductions from baseline up to –1.67% at 14 weeks,^40,41 up to –1.1% at 26 weeks,^23,26 and up to –1.14% at 52 weeks.²⁵ The reductions quoted generally refer to means, and individual patients may have greater or lesser responses. Also, baseline HbA1c is a significant determinant of the potential HbA1c reduction. Higher baseline values drop more significantly than do baseline values that are closer to normal.

Weight reduction with GLP-1 receptor agonists. In addition to effective glucose lowering, the GLP-1 receptor agonists, particularly exendin-4 agonists, produced beneficial effects on weight (Table 1). Exenatide BID elicited mean weight reductions up to –3.6 kg at 30 weeks^21,31,32 and –5.3 kg at 3.5 years.³⁷ Exenatide once weekly resulted in mean weight reductions of up to –3.8 kg at 15 weeks²² and –3.7 kg at 30 weeks.²¹ Effects on weight with liraglutide varied from a mean reduction of up to –2.99 kg to a slight gain of up to +0.13 kg at 14 weeks^40,41 and with weight loss of up to –2.8 kg at 26 weeks^23,26 and up to –2.5 kg at 52 weeks.²⁵ In this review, only exenatide has been assessed in insulin-comparator studies, where it was shown to reduce weight compared with the insulin analogues, which led to weight gain.^34–36

Hypoglycemia. Patients receiving exenatide experienced lower rates of hypoglycemia (up to 17%) than patients treated with either insulin glargine or insulin aspart (~25%).^34,36 The rate of hypoglycemia with exenatide is comparable to that seen with metformin (up to 21%) in a systematic review of oral antidiabetes agents conducted by the Agency for Healthcare Research and Quality.⁶² No major hypoglycemic events were reported in the liraglutide studies reviewed. The incidence of hypoglycemia reported with DPP-4 inhibitors (Table 2) is also low (2% or less in most studies). The glucose-dependent mechanisms of the incretin-based therapies minimizes the risk of hypoglycemia.

DPP-4 inhibitors and sustained HbA1c reduction. The effects of the DPP-4 inhibitors on HbA1c and weight, either as monotherapy or in combination with other agents, were evaluated in studies ranging in duration from 12 to 52 weeks (Table 2). No studies were identified that compared the glycemic control effects of DPP-4 inhibitors and insulin analogues. Sitagliptin led to a mean reduction in HbA1c from baseline of up to –0.65% at 12 weeks,^43,45 up to –0.48% at 18 weeks,⁴⁴ up to –0.85% at 24 weeks,^42,46,47,50 up to –1.0% at 30 weeks,⁴⁹ and up to –0.67% at 52 weeks.⁴⁸ Saxagliptin mean reductions in HbA1c ranged from –0.43% to –1.17%.^51–54 Data from four 24-week T2DM studies^56–60 showed vildagliptin reducing HbA1c up to –1.4% at 24 weeks, with the greatest reduction in a study that involved drug-naïve patients with a relatively short duration of disease (mean, 1.2 years).⁵⁹ Reductions in HbA1c of –1.0% were sustained in a 52-week study⁶¹ and its 52-week extension.⁵⁸

DPP-4 inhibitors: weight neutral. The DPP-4 inhibitors appear to have a weight-neutral effect (Table 2). The effects of sitagliptin on weight ranged from a loss of –1.5 kg⁴⁸ at 52 weeks to a gain of +1.8 kg at 24 weeks.⁵⁰ Weight changes with saxagliptin ranged from a mean reduction of –1.8 kg⁵³ to a gain of +0.7 kg.⁵¹ Two vildagliptin studies showed varying effects on weight ranging from a loss of up to –1.8 kg from baseline⁵⁶ to a gain of up to +1.3 kg⁵⁷ relative to placebo, both at 24 weeks.

Potential for CV risk reduction

Potentially beneficial effects on CV risk factors, including blood pressure (ie, reduction) and lipid concentrations (ie, differential effects on low-density lipoprotein and high-density lipoprotein cholesterol), were identified in seven GLP-1 receptor studies—three with exenatide (two with exenatide BID,^37,38 and one with the investigational exenatide once weekly²¹) and four with liraglutide.^23,25,26,41 For the DPP-4 inhibitors, three studies were identified—two with sitagliptin^45,50 and one with vildagliptin⁶¹—in which potentially beneficial effects on CV risk factors were demonstrated.The data have been encouraging, although the clinical implications have yet to be fully understood.

Head-to-head comparison

A recent study compared the effects of the GLP-1 receptor agonist exenatide and the DPP-4 inhibitor sitagliptin on postprandial glucose (PPG) concentrations, insulin and glucagon secretion, gastric intake, and caloric intake.³⁹ Although limited by a short treatment duration (2 weeks), the study showed that the GLP-1 receptor agonist had a greater effect than the DPP-4 inhibitor in reducing PPG concentrations, a more potent effect in increasing insulin secretion and decreasing postprandial glucagon secretion, and a relatively greater effect in reducing caloric intake; and that it decreased the rate of gastric emptying (sitagliptin had no effect). These differences suggest that exenatide may provide a greater degree of GLP-1 receptor activation than the more physiologic concentrations of GLP-1 reached with DPP-4 inhibition.³⁹ Results of a scintigraphic study showed that exenatide substantially slows the gastric emptying that is accelerated in patients with T2DM. This could be another beneficial mechanism in treating postprandial glycemia.⁶³

Adverse effects

Exenatide has shown effects on hepatic injury markers (ie, improvement in alanine and aspartate aminotransferases) for up to 3.5 years of treatment.³⁷ For the GLP-1 receptor agonist and DPP-4 inhibitor studies reviewed, the adverse events were generally mild and included nausea and vomiting, nasopharyngitis, and mild hypoglycemia.

 

 

Meta-analysis conclusions

The published clinical trial data presented in this review expand the body of evidence on the safety and efficacy of incretin-based therapy in patients with T2DM. These data include the results of a meta-analysis by Amori et al,¹⁷ which examined randomized controlled trials of 12 weeks’ or longer duration that compared incretin-based therapy with placebo or other diabetes medications and reported HbA1c changes in adults with T2DM. The meta-analysis showed that incretin-based therapies reduced HbA1c more than placebo (weighted mean difference, –0.97% [95% confidence interval (CI), –1.13% to –0.81%] for GLP-1 receptor agonists and –0.74% [95% CI, –0.85% to –0.62%] for DPP-4 inhibitors) and were noninferior to other antidiabetes agents. Treatment with a GLP-1 receptor agonist (ie, exenatide) caused weight loss (–1.4 kg and –4.8 kg vs placebo and insulin, respectively) while DPP-4 inhibitors (ie, sitagliptin, vildagliptin) were weight neutral.¹⁷

Beta-cell function

Evidence regarding the effects of incretin-based therapies, particularly the exendin-4 GLP-1 receptor agonists, on beta-cell function in patients with T2DM continues to accumulate. When assessing long-term (1 year) exenatide treatment in patients with T2DM, a trial (n = 69) comparing exenatide with the basal insulin analogue insulin glargine showed that exenatide and insulin glargine resulted in similar reductions in HbA1c (–0.8% vs –0.7%; P = .55).⁶⁴ However, exenatide significantly reduced body weight while insulin glargine resulted in weight gain (–3.6 kg vs +1.0 kg; P < .0001). In terms of beta-cell function, arginine-stimulated C-peptide secretion during hyperglycemia increased 2.46-fold from baseline after 52 weeks of exenatide treatment compared with 1.31-fold with insulin glargine treatment (P < .0001).⁶⁴

With respect to the direct beta-cell effects of liraglutide, a preclinical study reported that liraglutide improved glucose homeostasis in marginal mass islet transplantation in diabetic mice.⁶⁵ In this study, liraglutide was shown, in a mouse model, to reduce the time to normoglycemia after islet cell transplantation (median time, 1 vs 72.5 days; P < .0001). The effects of liraglutide on beta-cell function also were assessed in 13 patients with T2DM. After 7 days of treatment, liraglutide improved beta-cell function, which was associated with improvement in glucose concentration.⁶⁶ Liraglutide improved potentiation of insulin secretion during the first meal, owing in part to restoration of the potentiation peak (which is markedly blunted in T2DM), in a phenomenon similar to that observed with exenatide.⁶⁷

Beneficial effects on beta-cell function have also been reported with DPP-4 inhibitors. In a model-based analysis of patients with T2DM, it was shown that sitagliptin improved basal, static, and dynamic responsiveness of pancreatic beta cells to glucose. The results were observed when sitagliptin was administered both as an add-on to metformin therapy and as monotherapy.⁶⁸ A 52-week, double-blind, randomized, parallel-group study compared vildagliptin 50 mg/day and placebo in 306 patients with T2DM and mild hyperglycemia (HbA1c, 6.2% to 7.5%). Vildagliptin was shown to significantly increase fasting insulin secretory tone, glucose sensitivity, and rate sensitivity, all of which are aspects of beta-cell function.⁶⁹

Summary

Based on the ability of incretin-based therapies to address various disease mechanisms, including beta-cell defects (ie, hyperglycemia), hormone-related abnormalities (ie, hyperglucagonemia, incretin deficiency/resistance), and accelerated gastric emptying (especially with GLP-1 receptor agonists); their favorable effects on weight (reduction with GLP-1 receptor agonists and neutral with DPP-4 inhibitors); their beneficial effects on CV risk factors; and their good safety profile (ie, hypoglycemia risk comparable with metformin), these agents could be considered therapeutic advances for the treatment of patients with T2DM.

INCRETIN-BASED THERAPIES IN GUIDELINES AND ALGORITHMS

The 2007 AACE medical guidelines for clinical practice for the management of diabetes recognized the place of the incretin-based therapies and included them among the pharmacologic options.⁵ Exenatide was specifically recommended for combination therapy with metformin, a sulfonylurea (secretagogue), a sulfonylurea plus metformin, or a TZD. Sitagliptin was recommended for use as monotherapy or in combination with metformin or a TZD.⁵

In 2009, the American Diabetes Association (ADA) and the European Association for the Study of Diabetes convened a consensus panel to produce an algorithm for the initiation and adjustment of therapy for patients with T2DM. In this algorithm, GLP-1 receptor agonists were considered appropriate in certain clinical scenarios (eg, when hypoglycemia was an issue or weight loss was a major consideration during treatment). However, the groups also noted a need for more data on long-term safety and the cost of treatment with incretin-based therapies.⁷⁰

The AACE and the American College of Endocrinology recently developed “road maps” for managing patients with T2DM. In patients with T2DM who are naïve to therapy, DPP-4 inhibitors are among the recommended first options when the initial HbA1c is 6.0% to 7.0% and as a combination therapy component when HbA1c reaches 7.0% to 9.0%. In patients who have already received monotherapy for 2 to 3 months and whose HbA1c is 6.5% to 8.5%, treatment options include combination therapy with a DPP-4 inhibitor and metformin or a TZD. Another option includes the initiation of treatment with a GLP-1 receptor agonist in combination with a TZD, with metformin or a sulfonylurea, or with metformin and a sulfonylurea.⁷¹

The role of GLP-1 receptor agonist therapies and their incorporation into T2DM treatment algorithms was noted at the 2008 annual meeting of the ADA. In the Banting lecture, Ralph A. DeFronzo, MD, advocated the early use of triple-drug therapy with metformin, exenatide, and a TZD in the management of patients with T2DM.⁹

CONCLUSION

T2DM, which is linked to weight gain and obesity, is a complex disease that predisposes patients to and is associated with CVD. A better understanding and appreciation of the role of the incretin system in the pathogenesis of T2DM has led to the development of incretin-based therapies, such as the GLP-1 receptor agonists and DPP-4 inhibitors. As more experimental and clinical evidence becomes available, subtle nuances are emerging that distinguish the roles of these two therapeutic classes.

The prevalence of type 2 diabetes mellitus (T2DM) is increasing exponentially worldwide. According to the Centers for Disease Control and Prevention, more than 23 million Americans had diabetes in 2007.¹ Globally, the prevalence of diabetes, of which T2DM accounts for 90% to 95% of cases,¹ is expected to increase from 171 million in 2000 to 366 million in 2030.² The National Health and Nutrition Examination Survey (NHANES) showed that about 66% of Americans were overweight or obese between 2003–2004.³ Data from a Swedish National Diabetes Register study showed both overweight and obesity as independent risk factors for cardiovascular disease (CVD) in patients with T2DM.⁴

This article presents an overview of the evolving concepts of the pathophysiology of T2DM, with a focus on two new therapeutic classes: the glucagon-like peptide–1 (GLP-1) receptor agonists and the dipeptidyl peptidase–4 (DPP-4) inhibitors.

THE PATHOPHYSIOLOGY OF T2DM

The American Association of Clinical Endocrinologists (AACE) describes T2DM as “a progressive, complex metabolic disorder characterized by coexisting defects of multiple organ sites including insulin resistance in muscle and adipose tissue, a progressive decline in pancreatic insulin secretion, unrestrained hepatic glucose production, and other hormonal deficiencies.”⁵ Other defects include accelerated gastric emptying in patients with T2DM, especially those who are obese or who have the disease for a long duration.^6,7

Hormonal deficiencies in T2DM are related to abnormalities in the secretion of the beta-cell hormone amylin, the alpha-cell hormone glucagon, and the incretin hormones GLP-1 and glucose-dependent insulinotropic polypeptide (GIP).^8,9 In addition to the triumvirate of core defects associated with T2DM (involvement of the pancreatic beta cell, muscle, and liver), other mechanisms of disease onset have been advanced, including accelerated lipolysis, hyperglucagonemia, and incretin deficiency/resistance.⁹ Also, the rate of basal hepatic glucose production is markedly increased in patients with T2DM, which is closely correlated with elevations in fasting plasma glucagon concentration.⁹

The incretin effect—the intestinal augmentation of secretion of insulin—attributed to GLP-1 and GIP is reduced in patients with T2DM.¹⁰ The secretion of GIP may be normal or elevated in patients with T2DM while the secretion of GLP-1 is deficient; however, cellular responsiveness to GLP-1 is preserved while responsiveness to GIP is diminished.¹¹

Both endogenous and exogenous GLP-1 and GIP are degraded in vivo and in vitro by the enzyme DPP-4,¹²
a ubiquitous, membrane-spanning, cell-surface aminopeptidase that preferentially cleaves peptides with a proline or alanine residue in the second amino-terminal position. DPP-4 is widely expressed (eg, in the liver, lungs, kidney, lymphocytes, epithelial cells, endothelial cells). The role of DPP-4 in the immune system stems from its exopeptidase activity and its interactions with various molecules, including cyto­kines and chemokines.¹³

INCRETIN-BASED THERAPIES: GLP-1 RECEPTOR AGONISTS AND DPP-4 INHIBITORS

Exenatide is a GLP-1 receptor agonist that is resistant to DPP-4 degradation. Based on preclinical studies, exenatide, which shares a 53% amino acid sequence identity with human GLP-1, is approximately 5,500 times more potent than endogenous GLP-1 in glucose lowering.^14,15 Among the acute actions of exenatide is glucose-dependent insulinotropism, the end result of which may be a reduced risk of hypoglycemia.¹⁶ This contrasts with insulin secretagogues (eg, sulfonylureas), which increase insulin secretion regardless of glucose concentrations.

Exenatide received US Food and Drug Administration (FDA) approval in 2005 and is indicated for the treatment of patients with T2DM.^13,17 Exenatide is administered BID as a subcutaneous (SC) injection in doses of 5 or 10 μg within 1 hour before the two major meals of the day, which should be eaten about 6 hours apart.¹⁸

Approved in 2006, sitagliptin was the first DPP-4 inhibitor indicated for adjunctive therapy to lifestyle modifications for the treatment of patients with T2DM.¹⁷ The recommended dosage of oral sitagliptin is 100 mg QD. A single-tablet formulation of the combination of sitagliptin and metformin was approved by the FDA in 2007.¹⁹ Another DPP-4 inhibitor, saxagliptin, was approved in July 2009 for treatment of patients with T2DM either as monotherapy or in combination with metformin, sulfonylurea, or a thiazolidinedione (TZD).²⁰ The DPP-4 inhibitor vildagliptin is approved in the European Union and Latin America but not in the United States. Vildagliptin is available as a 50- or 100-mg daily dosage; it has been recommended for use at 50 mg QD in combination with a sulfonylurea or at 50 mg BID with either metformin or a TZD.¹⁸

GLP-1 RECEPTOR AGONISTS AND DPP-4 INHIBITORS IN DEVELOPMENT

Exenatide is currently being evaluated as a once-weekly formulation.^21,22 Compared with the BID formulation, exenatide once weekly has been shown to produce significantly greater improvements in glycemic control, with similar reductions in body weight and no increased risk of hypoglycemia.²¹

Also undergoing regulatory review is the partly DPP-4–resistant acylated GLP-1 receptor agonist liraglutide.¹³ Liraglutide, a human analogue GLP-1 receptor agonist, has 97% linear amino acid sequence homology to human GLP-1.^23,24 Based on its prolonged degradation time and resulting 10- to 14-hour half-life, liraglutide is anticipated to be dosed once daily.^13,25,26

Other GLP-1 receptor agonists and DPP-4 inhibitors are in varying stages of development.²⁷ Albiglutide is a long-acting GLP-1 receptor agonist that is generated by the genetic fusion of a DPP-4–resistant GLP-1 to human albumin. Based on pharmacokinetic studies, albiglutide has a half-life of 6 to 8 days. AVE0010, an exendin-4-based GLP-1 receptor agonist, was shown in a 28-day T2DM clinical trial to have an affinity four times greater than native GLP-1 for the human GLP-1 receptor.²⁷ Taspoglutide (R1583), a human analogue GLP-1 receptor agonist, was evaluated in three randomized, placebo-controlled studies as a GLP-1 receptor agonist. Alogliptin, a DPP-4 inhibitor currently in development, has been shown to be safe and effective in studies as monotherapy and in combination with other antidiabetes agents.^28–30

 

 

CLINICAL TRIALS: GLP-1 RECEPTOR AGONISTS AND DPP-4 INHIBITORS

This section summarizes clinical trials of GLP-1 receptor agonists and DPP-4 inhibitors. The summary is based on literature published from 2005 to 2009 relevant to phase 3 or 4 T2DM clinical trials with currently available agents, or agents with pending new drug applications.

Table 1 summarizes the data on the effects of the GLP-1 receptor agonists on glucose lowering based on glycosylated hemoglobin (HbA1c) mean changes from baseline, body weight, and hypoglycemia. Eleven studies were identified for exenatide, including three pivotal trials,^31–33 three insulin-comparator studies,^34–36 one long-term study,³⁷ one monotherapy study (a use for which it is not currently indicated),³⁸ one head-to-head study with a DPP-4 inhibitor,³⁹ and two studies with exenatide once weekly (which is currently investigational).^21,22 Five primary efficacy studies with liraglutide were also identified.^{23,25,26,40,41}

Table 2 summarizes the corresponding data for the DPP-4 inhibitors. Ten studies with sitagliptin were identified, including four monotherapy studies,^42–45 one head-to-head study with a GLP-1 receptor agonist,³⁹ and five studies in which sitagliptin was used in combination or as add-on therapy.^46–50 Five saxagliptin studies are reviewed, including two in which saxagliptin was used in combination with metformin and one in combination with glyburide.^51–55 Six studies with vildagliptin were reviewed,^56–61 but no trials specific to the single-tablet formulation of sitagliptin plus metformin were identified.

Effects on HbA1c and weight

GLP-1 receptor agonists reduced HbA1c. Based on the studies reviewed in Table 1, exenatide BID reduced baseline HbA1c by a maximum of –1.5% at 30 weeks.^21,31,32 Exenatide has demonstrated sustained reductions in HbA1c of –0.8% for up to 3.5 years in an open-label extension trial.³⁷ Even greater reductions in HbA1c (–1.4% at 15 weeks and –1.9% at 30 weeks) have been reported with the once-weekly formulation under clinical development.^21,22 Liraglutide, another GLP-1 receptor agonist under development, has reported HbA1c reductions from baseline up to –1.67% at 14 weeks,^40,41 up to –1.1% at 26 weeks,^23,26 and up to –1.14% at 52 weeks.²⁵ The reductions quoted generally refer to means, and individual patients may have greater or lesser responses. Also, baseline HbA1c is a significant determinant of the potential HbA1c reduction. Higher baseline values drop more significantly than do baseline values that are closer to normal.

Weight reduction with GLP-1 receptor agonists. In addition to effective glucose lowering, the GLP-1 receptor agonists, particularly exendin-4 agonists, produced beneficial effects on weight (Table 1). Exenatide BID elicited mean weight reductions up to –3.6 kg at 30 weeks^21,31,32 and –5.3 kg at 3.5 years.³⁷ Exenatide once weekly resulted in mean weight reductions of up to –3.8 kg at 15 weeks²² and –3.7 kg at 30 weeks.²¹ Effects on weight with liraglutide varied from a mean reduction of up to –2.99 kg to a slight gain of up to +0.13 kg at 14 weeks^40,41 and with weight loss of up to –2.8 kg at 26 weeks^23,26 and up to –2.5 kg at 52 weeks.²⁵ In this review, only exenatide has been assessed in insulin-comparator studies, where it was shown to reduce weight compared with the insulin analogues, which led to weight gain.^34–36

Hypoglycemia. Patients receiving exenatide experienced lower rates of hypoglycemia (up to 17%) than patients treated with either insulin glargine or insulin aspart (~25%).^34,36 The rate of hypoglycemia with exenatide is comparable to that seen with metformin (up to 21%) in a systematic review of oral antidiabetes agents conducted by the Agency for Healthcare Research and Quality.⁶² No major hypoglycemic events were reported in the liraglutide studies reviewed. The incidence of hypoglycemia reported with DPP-4 inhibitors (Table 2) is also low (2% or less in most studies). The glucose-dependent mechanisms of the incretin-based therapies minimizes the risk of hypoglycemia.

DPP-4 inhibitors and sustained HbA1c reduction. The effects of the DPP-4 inhibitors on HbA1c and weight, either as monotherapy or in combination with other agents, were evaluated in studies ranging in duration from 12 to 52 weeks (Table 2). No studies were identified that compared the glycemic control effects of DPP-4 inhibitors and insulin analogues. Sitagliptin led to a mean reduction in HbA1c from baseline of up to –0.65% at 12 weeks,^43,45 up to –0.48% at 18 weeks,⁴⁴ up to –0.85% at 24 weeks,^42,46,47,50 up to –1.0% at 30 weeks,⁴⁹ and up to –0.67% at 52 weeks.⁴⁸ Saxagliptin mean reductions in HbA1c ranged from –0.43% to –1.17%.^51–54 Data from four 24-week T2DM studies^56–60 showed vildagliptin reducing HbA1c up to –1.4% at 24 weeks, with the greatest reduction in a study that involved drug-naïve patients with a relatively short duration of disease (mean, 1.2 years).⁵⁹ Reductions in HbA1c of –1.0% were sustained in a 52-week study⁶¹ and its 52-week extension.⁵⁸

DPP-4 inhibitors: weight neutral. The DPP-4 inhibitors appear to have a weight-neutral effect (Table 2). The effects of sitagliptin on weight ranged from a loss of –1.5 kg⁴⁸ at 52 weeks to a gain of +1.8 kg at 24 weeks.⁵⁰ Weight changes with saxagliptin ranged from a mean reduction of –1.8 kg⁵³ to a gain of +0.7 kg.⁵¹ Two vildagliptin studies showed varying effects on weight ranging from a loss of up to –1.8 kg from baseline⁵⁶ to a gain of up to +1.3 kg⁵⁷ relative to placebo, both at 24 weeks.

Potential for CV risk reduction

Potentially beneficial effects on CV risk factors, including blood pressure (ie, reduction) and lipid concentrations (ie, differential effects on low-density lipoprotein and high-density lipoprotein cholesterol), were identified in seven GLP-1 receptor studies—three with exenatide (two with exenatide BID,^37,38 and one with the investigational exenatide once weekly²¹) and four with liraglutide.^23,25,26,41 For the DPP-4 inhibitors, three studies were identified—two with sitagliptin^45,50 and one with vildagliptin⁶¹—in which potentially beneficial effects on CV risk factors were demonstrated.The data have been encouraging, although the clinical implications have yet to be fully understood.

Head-to-head comparison

A recent study compared the effects of the GLP-1 receptor agonist exenatide and the DPP-4 inhibitor sitagliptin on postprandial glucose (PPG) concentrations, insulin and glucagon secretion, gastric intake, and caloric intake.³⁹ Although limited by a short treatment duration (2 weeks), the study showed that the GLP-1 receptor agonist had a greater effect than the DPP-4 inhibitor in reducing PPG concentrations, a more potent effect in increasing insulin secretion and decreasing postprandial glucagon secretion, and a relatively greater effect in reducing caloric intake; and that it decreased the rate of gastric emptying (sitagliptin had no effect). These differences suggest that exenatide may provide a greater degree of GLP-1 receptor activation than the more physiologic concentrations of GLP-1 reached with DPP-4 inhibition.³⁹ Results of a scintigraphic study showed that exenatide substantially slows the gastric emptying that is accelerated in patients with T2DM. This could be another beneficial mechanism in treating postprandial glycemia.⁶³

Adverse effects

Exenatide has shown effects on hepatic injury markers (ie, improvement in alanine and aspartate aminotransferases) for up to 3.5 years of treatment.³⁷ For the GLP-1 receptor agonist and DPP-4 inhibitor studies reviewed, the adverse events were generally mild and included nausea and vomiting, nasopharyngitis, and mild hypoglycemia.

 

 

Meta-analysis conclusions

The published clinical trial data presented in this review expand the body of evidence on the safety and efficacy of incretin-based therapy in patients with T2DM. These data include the results of a meta-analysis by Amori et al,¹⁷ which examined randomized controlled trials of 12 weeks’ or longer duration that compared incretin-based therapy with placebo or other diabetes medications and reported HbA1c changes in adults with T2DM. The meta-analysis showed that incretin-based therapies reduced HbA1c more than placebo (weighted mean difference, –0.97% [95% confidence interval (CI), –1.13% to –0.81%] for GLP-1 receptor agonists and –0.74% [95% CI, –0.85% to –0.62%] for DPP-4 inhibitors) and were noninferior to other antidiabetes agents. Treatment with a GLP-1 receptor agonist (ie, exenatide) caused weight loss (–1.4 kg and –4.8 kg vs placebo and insulin, respectively) while DPP-4 inhibitors (ie, sitagliptin, vildagliptin) were weight neutral.¹⁷

Beta-cell function

Evidence regarding the effects of incretin-based therapies, particularly the exendin-4 GLP-1 receptor agonists, on beta-cell function in patients with T2DM continues to accumulate. When assessing long-term (1 year) exenatide treatment in patients with T2DM, a trial (n = 69) comparing exenatide with the basal insulin analogue insulin glargine showed that exenatide and insulin glargine resulted in similar reductions in HbA1c (–0.8% vs –0.7%; P = .55).⁶⁴ However, exenatide significantly reduced body weight while insulin glargine resulted in weight gain (–3.6 kg vs +1.0 kg; P < .0001). In terms of beta-cell function, arginine-stimulated C-peptide secretion during hyperglycemia increased 2.46-fold from baseline after 52 weeks of exenatide treatment compared with 1.31-fold with insulin glargine treatment (P < .0001).⁶⁴

With respect to the direct beta-cell effects of liraglutide, a preclinical study reported that liraglutide improved glucose homeostasis in marginal mass islet transplantation in diabetic mice.⁶⁵ In this study, liraglutide was shown, in a mouse model, to reduce the time to normoglycemia after islet cell transplantation (median time, 1 vs 72.5 days; P < .0001). The effects of liraglutide on beta-cell function also were assessed in 13 patients with T2DM. After 7 days of treatment, liraglutide improved beta-cell function, which was associated with improvement in glucose concentration.⁶⁶ Liraglutide improved potentiation of insulin secretion during the first meal, owing in part to restoration of the potentiation peak (which is markedly blunted in T2DM), in a phenomenon similar to that observed with exenatide.⁶⁷

Beneficial effects on beta-cell function have also been reported with DPP-4 inhibitors. In a model-based analysis of patients with T2DM, it was shown that sitagliptin improved basal, static, and dynamic responsiveness of pancreatic beta cells to glucose. The results were observed when sitagliptin was administered both as an add-on to metformin therapy and as monotherapy.⁶⁸ A 52-week, double-blind, randomized, parallel-group study compared vildagliptin 50 mg/day and placebo in 306 patients with T2DM and mild hyperglycemia (HbA1c, 6.2% to 7.5%). Vildagliptin was shown to significantly increase fasting insulin secretory tone, glucose sensitivity, and rate sensitivity, all of which are aspects of beta-cell function.⁶⁹

Summary

Based on the ability of incretin-based therapies to address various disease mechanisms, including beta-cell defects (ie, hyperglycemia), hormone-related abnormalities (ie, hyperglucagonemia, incretin deficiency/resistance), and accelerated gastric emptying (especially with GLP-1 receptor agonists); their favorable effects on weight (reduction with GLP-1 receptor agonists and neutral with DPP-4 inhibitors); their beneficial effects on CV risk factors; and their good safety profile (ie, hypoglycemia risk comparable with metformin), these agents could be considered therapeutic advances for the treatment of patients with T2DM.

INCRETIN-BASED THERAPIES IN GUIDELINES AND ALGORITHMS

The 2007 AACE medical guidelines for clinical practice for the management of diabetes recognized the place of the incretin-based therapies and included them among the pharmacologic options.⁵ Exenatide was specifically recommended for combination therapy with metformin, a sulfonylurea (secretagogue), a sulfonylurea plus metformin, or a TZD. Sitagliptin was recommended for use as monotherapy or in combination with metformin or a TZD.⁵

In 2009, the American Diabetes Association (ADA) and the European Association for the Study of Diabetes convened a consensus panel to produce an algorithm for the initiation and adjustment of therapy for patients with T2DM. In this algorithm, GLP-1 receptor agonists were considered appropriate in certain clinical scenarios (eg, when hypoglycemia was an issue or weight loss was a major consideration during treatment). However, the groups also noted a need for more data on long-term safety and the cost of treatment with incretin-based therapies.⁷⁰

The AACE and the American College of Endocrinology recently developed “road maps” for managing patients with T2DM. In patients with T2DM who are naïve to therapy, DPP-4 inhibitors are among the recommended first options when the initial HbA1c is 6.0% to 7.0% and as a combination therapy component when HbA1c reaches 7.0% to 9.0%. In patients who have already received monotherapy for 2 to 3 months and whose HbA1c is 6.5% to 8.5%, treatment options include combination therapy with a DPP-4 inhibitor and metformin or a TZD. Another option includes the initiation of treatment with a GLP-1 receptor agonist in combination with a TZD, with metformin or a sulfonylurea, or with metformin and a sulfonylurea.⁷¹

The role of GLP-1 receptor agonist therapies and their incorporation into T2DM treatment algorithms was noted at the 2008 annual meeting of the ADA. In the Banting lecture, Ralph A. DeFronzo, MD, advocated the early use of triple-drug therapy with metformin, exenatide, and a TZD in the management of patients with T2DM.⁹

CONCLUSION

T2DM, which is linked to weight gain and obesity, is a complex disease that predisposes patients to and is associated with CVD. A better understanding and appreciation of the role of the incretin system in the pathogenesis of T2DM has led to the development of incretin-based therapies, such as the GLP-1 receptor agonists and DPP-4 inhibitors. As more experimental and clinical evidence becomes available, subtle nuances are emerging that distinguish the roles of these two therapeutic classes.

According to the American Association of Clinical Endocrinologists (AACE) and the American Diabetes Association (ADA), glycosylated hemoglobin (HbA1c) in patients with diabetes should be maintained at 6.5% or less (AACE) or at less than 7.0% (ADA). Both organizations support an aggressive stepwise approach that includes medication and lifestyle modification, with strategies and clinical attention devoted to avoiding significant hypoglycemia.^1,2 Yet, despite the introduction of new antidiabetes agents, most current management strategies are offset by limitations in achieving and maintaining glycemic targets needed to provide optimal care for patients with diabetes, more than 90% of whom have type 2 diabetes mellitus (T2DM).^3,4

Nationally, glycemic control among patients with T2DM has improved but is still far from optimal. According to data from the 1999–2000 National Health and Nutrition Examination Survey (NHANES), glycemic control (HbA1c < 7.0%) rates were 35.8% for patients with T2DM.⁵ In a more recent report (NHANES 1999–2004), fewer than half (48.4%) of adult patients with diagnosed diabetes achieved HbA1c levels below 7.0%.^5,6 Factors contributing to these data include earlier onset and earlier detection of T2DM.⁷

CHANGING TREATMENT TRENDS

Available treatments for patients with T2DM include secretagogues, such as sulfonylureas and “glinides” (repaglinide and nateglinide), metformin, thiazolidinediones (TZDs), and dipeptidyl peptidase–4 (DPP-4) inhibitors among oral medications, and insulin and glucagon-like peptide–1 (GLP-1) receptor agonists among parenterally administered agents. According to the latest published data on prescribing patterns for patients with T2DM, analyses of the National Disease and Therapeutic Index (1994–2007) and the National Prescription Audit (2001–2007), sulfonylurea use decreased from 67% of treatment visits in 1994 to 34% of visits in 2007.⁸ By 2007, metformin, used in 54% of treatment visits, and TZDs, used in 28%, were the most frequently administered antidiabetes agents. Insulin use declined from 38% of visits during which a treatment was administered in 1994 to 25% of visits in 2000, but had increased subsequently to 28% of visits in 2007.

SIGNIFICANCE OF CARDIOVASCULAR RISK

Clinical research has suggested that focusing solely on improving glycemic control may be insufficient to reduce overall morbidity and mortality associated with diabetes. Specifically, data from recent studies, including the Action to Control Cardiovascular Risk in Diabetes (ACCORD), the Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE), and the Veterans Affairs Diabetes Trial (VADT), emphasized that lowering HbA1c below 7% in a high-risk population of individuals with T2DM did not improve cardiovascular (CV) outcomes.^9–11 The observations confirm that risk factors, including weight, blood pressure (BP), and lipid levels, are vitally important in reducing morbidity and mortality in this population. This perception is further underscored by the NHANES 1999–2004 data, which showed poor concurrent control of HbA1c, BP, and lipids; only 13.2% of patients with diagnosed diabetes achieved all three target goals simultaneously.⁶ Similarly, a nationwide survey in Norway showed that only 13% of patients with T2DM concurrently achieved goals for HbA1c, BP, and lipids.¹²

In the Danish Steno-2 Study, patients with T2DM and persistent microalbuminuria were treated with either intensive target-driven therapy using multiple drugs or conventional multifactorial treatment. Over a mean period of 13.3 years (7.8 years of treatment plus 5.5 years of follow-up), intensive multifactorial intervention to control multiple CV risk factors, including HbA1c, BP, and lipids, was associated with a lower risk of death from CV causes (hazard ratio [HR], 0.43; 95% confidence interval [CI], 0.19 to 0.94; P = .04) and a lower risk of CV events (HR, 0.41; 95% CI, 0.25 to 0.67; P < .001) than was conventional therapy.¹³

This article clarifies the redefinition of treatment success in patients with T2DM based on targeting the underlying physiologic defects of the disease.

T2DM, OVERWEIGHT/OBESITY, AND CV DISEASE: CLOSELY LINKED

The incidence and prevalence of T2DM, overweight/obesity, and CV disease (CVD) are increasing worldwide. It is estimated that the worldwide prevalence of diabetes will increase from 171 million in 2000 to 366 million by 2030¹⁴; T2DM increases the risk of morbidity and mortality from microvascular (eg, neuropathic, retinopathic, nephropathic) and macrovascular (eg, coronary, peripheral vascular disease) complications.¹⁵ According to a Michigan health maintenance organization study (N = 1,364), the median annual direct cost of medical care for Caucasian patients with T2DM who were diet controlled, had a body mass index (BMI) of 30 kg/m² or higher, and had no vascular complications was estimated to be $1,700 for men and $2,100 for women.¹⁶ The actual cost of care for patients with T2DM may be much higher, since most patients present with multiple CV risk factors in addition to being overweight.

NHANES data show that approximately two-thirds of Americans are either overweight or obese¹⁷; overweight/obesity affects about 80% of adults diagnosed with T2DM.¹⁸ Overweight or obesity can increase the risk for developing T2DM by more than 90-fold and, in women, it can increase the risk for developing coronary heart disease (CHD) by sixfold.¹⁹ The close link between T2DM and CVD is underscored further with recent data from the Framingham Heart Study, which showed a high lifetime risk of CVD in patients with diabetes, heightened further by obesity. During the 30-year study period, the lifetime risk of CVD in normal-weight people with diabetes was 78.6% in men and 54.8% in women; the risk increased to 86.9% in obese men with diabetes and to 78.8% in obese women with diabetes.²⁰ The NHANES data also showed that the prevalence of T2DM increased in the past decade and that patients are being diagnosed at a younger age, from a mean age of 52 years in 1988–1994 to 46 years in 1999–2000.⁷

 

 

BRIDGING THE GAP FROM PATHOPHYSIOLOGY TO UNMET NEEDS

The paradigm behind the pathophysiology of T2DM has shifted from its perception as a simple “dual-defect” disease (ie, deficiency in insulin secretion and peripheral tissue insulin resistance) to a multidimensional disorder.^1,21 This new model includes overweight/obesity, insulin resistance, qualitative and quantitative defects in insulin secretion, and dysregulation in the secretion of other hormones, including the beta-cell hormone amylin, the alpha-cell hormone glucagon, and the gastrointestinal incretin hormones GLP-1 and glucose-dependent insulinotropic polypeptide.^21–23

The major target of antidiabetes agents is glycemic control, assessed by a reduction in HbA1c, but their effects on other metabolic factors and their adverse effects differ with each agent (Table 1).³ Whereas metformin and alpha-glucosidase inhibitors may help normalize glycemia with weight-neutral effects, many other agents, including insulin and its analogues, the “glinides,” first- and second-generation sulfonylureas, and TZDs, are associated with weight gain.^23,24 In addition, the propensity to induce hypoglycemia differs among agents and clearly reflects the mechanism of action of each drug. The observed limitations of older therapies treating a progressive disease that is associated with a number of comorbid conditions supports the need for continued development of new antidiabetes agents.

CLINICAL GUIDELINES AND CV RISK FACTOR MANAGEMENT

The best strategy for managing T2DM is a comprehensive approach that addresses the fundamental core defects plus associated factors that contribute to increased CV risk. Several specialty groups have suggested guidelines and algorithms for the management of T2DM and its comorbidities. These guidelines, including the ADA standards of medical care, the AACE standards in tandem with the American College of Endocrinology guidelines, and the recent joint statement from the ADA and the European Association for the Study of Diabetes (EASD), acknowledge that the core defects of T2DM and the associated CV risk factors (eg, weight gain, obesity, hypertension, dyslipidemia) are important in developing optimal treatment strategies.^1–3 Medical nutrition guidelines advocate weight loss as a key initial step in managing T2DM and the comorbidities that lead to elevated CV risk.^25,26 The National Institutes of Health and the US Department of Health and Human Services/US Department of Agriculture advocate regular physical activity, dietary assessment, and periodic comorbidity and weight assessment for all people, not just those with T2DM or CVD.^26,27

Weight reduction

Evidence in support of effective lifestyle intervention was demonstrated in the Action for Health in Diabetes (Look AHEAD) study. After 1 year, patients with T2DM treated with intensive lifestyle intervention lost an average of 8.6% of their initial weight compared with 0.7% in patients treated only with diabetes support and education (P < 0.001). The intensive-intervention patients also had a significant drop in HbA1c (from 7.3% to 6.6%; P < 0.001) and were able to reduce their antidiabetes, antihypertensive, and lipid-lowering medications.²⁸ More recent data from the Look AHEAD study reported that overweight patients with T2DM enrolled in a weight management program experienced significant weight loss, improved physical fitness, reduced physical symptoms, and overall improvement in health-related quality of life.²⁹ Thus, weight reduction appears to be a key component in reducing CV risk and improving quality of life in most patients with T2DM.^28–30

Hypertension

Hypertension is a major risk factor for microvascular complications and CVD, and may be associated with, or be the underlying result of, nephropathy.² BP control is clearly important in reducing the morbidity and mortality associated with T2DM. The recommended BP goal in patients with T2DM is less than 130/80 mm Hg.^1,2

Hyperlipidemia

According to the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III [ATP III]), diabetes is considered a CHD risk equivalent because it confers a high risk of new CHD developing within 10 years.³¹ In addition to the NCEP–ATP III guidelines, the ADA and the AACE have set target levels for lipids in patients with diabetes, including T2DM.^1,2,31 All three organizations have defined 100 mg/dL as the target level for low-density lipoprotein.

HbA1c and lifestyle intervention

The American Heart Association and the ADA initiated a call to action for global risk assessment for CVD and diabetes.³² According to their joint scientific statement, lifestyle intervention should be reinforced at every physician visit, and HbA1c should be monitored every 3 months until it is less than 7.0% and then rechecked every 6 months. Adjustments in intervention should be made if the HbA1c level is 7.0% or higher.³ A recent joint statement from the ADA and the EASD revised an earlier treatment algorithm for the initiation of therapy in patients with T2DM; the revision includes incretin therapies (ie, GLP-1 receptor agonists) as a tier 2 option, especially in patients in whom hypoglycemia and weight gain are concerns (Figure 1).³

 

 

EVOLUTION OF ANTIDIABETES THERAPIES

Traditional antidiabetes agents used in the treatment of patients with T2DM have focused mainly on insulin secretion and insulin resistance, with treatment success defined as achieving HbA1c goals with a reduced incidence of hypoglycemia.²³ Secretagogues, such as sulfonylureas and glinides, stimulate the pancreas to release insulin. Insulin sensitizers, such as TZDs and metformin, enhance the action of insulin in muscle and fat^1,3,23 and lower hepatic glucose production. The alpha-glucosidase inhibitors alter carbohydrate absorption from the gastrointestinal tract.¹ The extent to which each agent achieves treatment success in terms of glucose lowering depends on several factors, including intrinsic attributes, duration of disease, and baseline glycemic control.³

Newer agents for the treatment of T2DM include the incretin-based therapies—GLP-1 receptor agonists and DPP-4 inhibitors—which influence mechanisms beyond increasing pancreatic insulin secretion and decreasing peripheral insulin resistance (Table 2).²² The GLP-1 signaling pathway has been leveraged by two distinct pharmacologic approaches. The first involves the use of synthetic peptides with glucoregulatory effects similar to those of endogenous GLP-1 (GLP-1 receptor agonists). The second involves the use of DPP-4 inhibitors, small molecules that inhibit the proteolytic activity of DPP-4, leading to enhanced endogenous GLP-1 concentrations.²²

GLP-1 receptor agonists

Exenatide effects. Although many agents are in development, to date exenatide is the only GLP-1 receptor agonist approved by the US Food and Drug Administration (FDA).^8,33 Exenatide is an exendin-4 GLP-1 receptor agonist with multiple glucoregulatory effects, including enhanced glucose-dependent insulin secretion, reduced glucagon secretion and food intake, and slowed gastric emptying.^22,34 Exenatide is detectable in the circulation for up to 10 hours following subcutaneous (SC) administration²² and has a greater potency in reducing plasma glucose than GLP-1 in preclinical studies.^35,36

By virtue of its beneficial effects on glycemic control, weight, BP, and lipids, exenatide addresses some of the components of the metabolic syndrome.^37–41 In pivotal 30-week studies, exenatide was associated with HbA1c reductions that ranged from –0.40% to –0.86% from baseline and decreases in body weight of approximately –1 kg to –3 kg from baseline, without severe hypoglycemia.^37–39 The percentage of patients who reached the ADA goal of HbA1c less than 7.0% at 30 weeks ranged from 24% to 34%. The addition of exenatide to TZD therapy in a 16-week study was associated with mean reductions in HbA1c of –0.98%, fasting plasma glucose (FPG) concentration of –1.69 mmol/L (–30.42 mg/dL), and body weight of –1.51 kg.⁴⁰

A posthoc analysis of an open-label extension study involving patients who completed the original 30-week placebo-controlled studies showed that 46% of patients who remained on exenatide achieved the ADA goal of HbA1c less than 7.0% at 3 years.⁴¹ Exenatide administered for up to 3.5 years was associated with sustained reductions in HbA1c of –1.0% (P < .0001) and body weight of –5.3 kg (P < .001). Pancreatic beta-cell function, assessed by homeostasis model assessment, improved, as did BP, triglyceride, high-density lipoprotein, low-density lipoprotein, and aspartate aminotransferase levels.⁴¹

Comparison with insulin analogues. Comparative studies have highlighted the contrasting effects of exenatide and insulin analogues (eg, insulin glargine and fixed-ratio insulin).^42–45 In a 26-week trial comparing exenatide with insulin glargine in subjects with T2DM, both agents resulted in similar decreases in HbA1c. Exenatide was also associated with a –2.3-kg weight reduction, whereas insulin glargine was associated with a +1.8-kg weight gain.⁴² Although rates of symptomatic hypoglycemia were similar, there were fewer cases of nocturnal hypoglycemia with exenatide (0.9 event/patient-year vs 2.4 events/patient-year with insulin).

In a 32-week study comparing exenatide BID with titrated insulin glargine QD, the HbA1c reductions for exenatide and insulin glargine were comparable. However, body weight decreased –4.2 kg over two 16-week treatment periods with exenatide, but increased +3.3 kg over the same periods with the basal insulin analogue.⁴³ The incidence of hypoglycemia was lower with exenatide than with insulin glargine (14.7% vs 25.2%), although the difference was not statistically significant.

In another study that compared exenatide with biphasic insulin aspart, patients who were treated with exenatide also lost weight while those who received the fast-acting insulin analogue gained weight (between-group difference, –5.4 kg). Patients treated with exenatide also demonstrated greater reductions in postprandial plasma glucose (PPG) excursions following their morning (P < .001), midday (P = .002), and evening meals (P < .001).⁴⁴ Overall, hypoglycemia rates were similar at study end between exenatide and insulin aspart (4.7 events/patient-year vs 5.6 events/patient-year). In all of these studies, significant gastrointestinal adverse events (nausea and vomiting) occurred more frequently with exenatide, and more patients withdrew from exenatide than from insulin.

Formulations in development. Other advances in GLP-1 receptor agonist therapy include novel formulations under clinical development, such as exenatide once weekly^36,46 and liraglutide, a human analogue GLP-1 receptor agonist formulated for once-daily administration.^47,48 In a 52-week study in patients with T2DM, liraglutide significantly reduced HbA1c; the 1.2-mg SC QD dosage reduced HBA1c by –0.84% (P = .0014) and the 1.8-mg SC QD dosage by –1.14% (P < .0001). In comparison, glimepiride 8 mg orally QD achieved a –0.51% reduction. Liraglutide was also associated with greater reductions in weight, hypoglycemia, and systolic BP than glimepiride.⁴⁷

A 26-week study compared liraglutide (0.6, 1.2, and 1.8 mg SC QD), placebo, and glimepiride 4 mg QD in combination with metformin 1 g BID. HbA1c was reduced significantly in all liraglutide groups compared with placebo (P < .0001). Mean HbA1c decreased –1.0% with liraglutide 1.2 mg and 1.8 mg and with glimepiride; it decreased –0.7% with liraglutide 0.6 mg; and it increased +0.1% with placebo. Body weight decreased –1.8 kg to –2.8 kg in all liraglutide groups but increased +1.0 kg in the glimepiride group (P < .0001). The incidence of minor hypoglycemia with liraglutide (~3%) was comparable to that observed with placebo but less than that with glimepiride (17%; P < .001).⁴⁸

A once-weekly long-acting release (LAR) formulation of exenatide submitted to the FDA for approval may provide enhanced glycemic and weight control, potentially improving patient acceptance and adherence.^36,46 In a 15-week study, exenatide once weekly produced significant reductions in HbA1c, FPG, PPG, and body weight. There were no withdrawals due to adverse events, and the formation of anti-exenatide antibodies was not predictive of therapeutic end point response or adverse safety outcome. Instances of hypoglycemia were mild and not dose related.³⁶ In a 30-week study comparing exenatide LAR once weekly with exenatide BID, patients given exenatide LAR once weekly had significantly greater HbA1c reductions than did patients given exenatide BID (–1.9% vs –1.5%; P = .0023). Treatment adherence was 98% with both exenatide regimens, and no episodes of major hypoglycemia occurred with either formulation regardless of background sulfonylurea use. Favorable effects on BP and lipid profile were observed with both exenatide regimens.⁴⁶

 

 

DPP-4 inhibitors

The DPP-4 inhibitors (commonly called gliptins) inhibit the proteolytic cleavage of circulating GLP-1 by binding to the DPP-4 enzyme, increasing the concentration of endogenous GLP-1 approximately two- to threefold.^49–51 These concentrations result in more prompt and appropriate secretion of insulin and suppression of glucagon in response to a carbohydrate-containing snack or meal, with the change in glucagon correlating linearly with improved glucose tolerance.⁵¹

DPP-4 inhibitors, which are given orally, include sitagliptin and saxagliptin (approved in the United States) and vildagliptin (not approved in the United States but used in the European Union and Latin America).^8,22,33,52 Sitagliptin can be used either as monotherapy or in combination with metformin or a TZD.^8,49–55 Recently, a single-tablet formulation of sitagliptin plus metformin was granted regulatory approval.⁸

When used alone or in combination with metformin or pioglitazone, sitagliptin has been associated with significant reductions in HbA1c (of ~0.5% to 0.6% when used alone, ~0.7% with metformin, and ~0.9% with pioglitazone [P < .001 vs placebo]), with hypoglycemia occurring in 1.3% or less of the population.⁵⁴ In an 18-week study in which patients with T2DM who were inadequately controlled with metformin monotherapy were randomized to receive add-on sitagliptin (100 mg QD), rosiglitazone (8 mg QD), or placebo, sitagliptin reduced HbA1c –0.73% (P < .001 vs placebo) and reduced body weight –0.4 kg, while rosiglitazone reduced HbA1c –0.79% and increased body weight +1.5 kg.⁵⁵

To evaluate the effectiveness of sitagliptin and metformin as initial therapy, a 54-week study was completed in 885 patients with T2DM and inadequate glycemic control (HbA1c 7.5–11%) on diet and exercise.⁵⁶ Patients were evaluated on monotherapy with either sitagliptin (100 mg QD) or metformin (1 g or 2 g QD), or on initial therapy with the two in combination (sitagliptin 100 mg + metformin 1 mg or 2 mg QD). At week 54, in the all-patients-treated analysis, mean changes in HbA1c from baseline were –1.8% with sitagliptin plus metformin 2 g QD, –1.4% with sitagliptin plus metformin 1 g QD, –1.3% with metformin 2 g QD monotherapy, –1.0% with metformin 1 g QD monotherapy, and –0.8% with sitagliptin 100 mg QD monotherapy.

All treatments improved measures of beta-cell function (eg, homeostasis model assessment [HOMA]-beta, proinsulin/insulin ratio). Mean body weight decreased from baseline in the combination and metformin monotherapy groups and was unchanged from baseline in the sitagliptin monotherapy group. The incidence of hypoglycemia was low (1%–3%) across treatment groups. The incidence of gastrointestinal adverse experiences was evaluated with the coadministration of sitagliptin and metformin and appeared similar to that observed with use of metformin as monotherapy.⁵⁶ Thus, this study suggested that an initial combination of a DPP-IV inhibitor with metformin can improve glycemic control and markers of beta-cell function in patients with T2DM.

Incretin-based therapies compared

Studies in both healthy individuals and in patients with T2DM have shown that oral DPP-4 inhibitors such as sitagliptin increase endogenous GLP-1 concentrations by about twofold compared with placebo.^22,50 The pharma­cologic concentration of subcutaneously administered exenatide available for activating the GLP-1 receptor is significantly greater than the increased endogenous GLP-1 concentrations achieved with sitagliptin. In a recent clinical study comparing exenatide and sitagliptin in patients with T2DM, the mean 2-hour plasma concentration for exenatide was 64 pM compared with the mean 2-hour postprandial GLP-1 concentration of 15 pM for sitagliptin (baseline GLP-1 concentration was 7.2 pM).⁵⁷ While both agents were shown to be effective, exenatide appeared to have had a greater effect than sitagliptin in increasing insulin secretion and reducing postprandial glucagon secretion, leading to significantly (P < 0.0001) greater reductions in PPG.⁵⁷

Sitagliptin has been minimally associated with nausea, whereas patients who take exenatide need to be informed of the risk of usually mild to moderate, but sometimes severe, nausea and vomiting that tends to decrease over time.

For a detailed comparison of the effects of GLP-1 receptor agonists and DPP-4 inhibitors on HbA1c, weight, and hypoglycemia, see “Advances in therapy for type 2 diabetes: GLP–1 receptor agonists and DPP–4 inhibitors.”

CONCLUSION

Despite advances in diagnosis and treatment, T2DM, overweight/obesity, CVD, and their complications remain major public health burdens worldwide. The concepts that explain the pathophysiology of T2DM include the contribution of various factors beyond insulin secretion and insulin resistance, such as the role of incretin hormones in disease progression. A comprehensive approach to managing patients with T2DM requires targeting the fundamental defects of the disease and its comorbidities. Newer agents, including incretin-based therapies such as GLP-1 receptor agonists and DPP-4 inhibitors, address the fundamental defects of T2DM. The definition of treatment success in the management of T2DM will be redefined as more data become available on agents that exert beneficial effects not only on glycemia but on parameters that may influence overall CV health, such as weight, BP, and lipid profiles.

According to the American Association of Clinical Endocrinologists (AACE) and the American Diabetes Association (ADA), glycosylated hemoglobin (HbA1c) in patients with diabetes should be maintained at 6.5% or less (AACE) or at less than 7.0% (ADA). Both organizations support an aggressive stepwise approach that includes medication and lifestyle modification, with strategies and clinical attention devoted to avoiding significant hypoglycemia.^1,2 Yet, despite the introduction of new antidiabetes agents, most current management strategies are offset by limitations in achieving and maintaining glycemic targets needed to provide optimal care for patients with diabetes, more than 90% of whom have type 2 diabetes mellitus (T2DM).^3,4

Nationally, glycemic control among patients with T2DM has improved but is still far from optimal. According to data from the 1999–2000 National Health and Nutrition Examination Survey (NHANES), glycemic control (HbA1c < 7.0%) rates were 35.8% for patients with T2DM.⁵ In a more recent report (NHANES 1999–2004), fewer than half (48.4%) of adult patients with diagnosed diabetes achieved HbA1c levels below 7.0%.^5,6 Factors contributing to these data include earlier onset and earlier detection of T2DM.⁷

CHANGING TREATMENT TRENDS

Available treatments for patients with T2DM include secretagogues, such as sulfonylureas and “glinides” (repaglinide and nateglinide), metformin, thiazolidinediones (TZDs), and dipeptidyl peptidase–4 (DPP-4) inhibitors among oral medications, and insulin and glucagon-like peptide–1 (GLP-1) receptor agonists among parenterally administered agents. According to the latest published data on prescribing patterns for patients with T2DM, analyses of the National Disease and Therapeutic Index (1994–2007) and the National Prescription Audit (2001–2007), sulfonylurea use decreased from 67% of treatment visits in 1994 to 34% of visits in 2007.⁸ By 2007, metformin, used in 54% of treatment visits, and TZDs, used in 28%, were the most frequently administered antidiabetes agents. Insulin use declined from 38% of visits during which a treatment was administered in 1994 to 25% of visits in 2000, but had increased subsequently to 28% of visits in 2007.

SIGNIFICANCE OF CARDIOVASCULAR RISK

Clinical research has suggested that focusing solely on improving glycemic control may be insufficient to reduce overall morbidity and mortality associated with diabetes. Specifically, data from recent studies, including the Action to Control Cardiovascular Risk in Diabetes (ACCORD), the Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE), and the Veterans Affairs Diabetes Trial (VADT), emphasized that lowering HbA1c below 7% in a high-risk population of individuals with T2DM did not improve cardiovascular (CV) outcomes.^9–11 The observations confirm that risk factors, including weight, blood pressure (BP), and lipid levels, are vitally important in reducing morbidity and mortality in this population. This perception is further underscored by the NHANES 1999–2004 data, which showed poor concurrent control of HbA1c, BP, and lipids; only 13.2% of patients with diagnosed diabetes achieved all three target goals simultaneously.⁶ Similarly, a nationwide survey in Norway showed that only 13% of patients with T2DM concurrently achieved goals for HbA1c, BP, and lipids.¹²

In the Danish Steno-2 Study, patients with T2DM and persistent microalbuminuria were treated with either intensive target-driven therapy using multiple drugs or conventional multifactorial treatment. Over a mean period of 13.3 years (7.8 years of treatment plus 5.5 years of follow-up), intensive multifactorial intervention to control multiple CV risk factors, including HbA1c, BP, and lipids, was associated with a lower risk of death from CV causes (hazard ratio [HR], 0.43; 95% confidence interval [CI], 0.19 to 0.94; P = .04) and a lower risk of CV events (HR, 0.41; 95% CI, 0.25 to 0.67; P < .001) than was conventional therapy.¹³

This article clarifies the redefinition of treatment success in patients with T2DM based on targeting the underlying physiologic defects of the disease.

T2DM, OVERWEIGHT/OBESITY, AND CV DISEASE: CLOSELY LINKED

The incidence and prevalence of T2DM, overweight/obesity, and CV disease (CVD) are increasing worldwide. It is estimated that the worldwide prevalence of diabetes will increase from 171 million in 2000 to 366 million by 2030¹⁴; T2DM increases the risk of morbidity and mortality from microvascular (eg, neuropathic, retinopathic, nephropathic) and macrovascular (eg, coronary, peripheral vascular disease) complications.¹⁵ According to a Michigan health maintenance organization study (N = 1,364), the median annual direct cost of medical care for Caucasian patients with T2DM who were diet controlled, had a body mass index (BMI) of 30 kg/m² or higher, and had no vascular complications was estimated to be $1,700 for men and $2,100 for women.¹⁶ The actual cost of care for patients with T2DM may be much higher, since most patients present with multiple CV risk factors in addition to being overweight.

NHANES data show that approximately two-thirds of Americans are either overweight or obese¹⁷; overweight/obesity affects about 80% of adults diagnosed with T2DM.¹⁸ Overweight or obesity can increase the risk for developing T2DM by more than 90-fold and, in women, it can increase the risk for developing coronary heart disease (CHD) by sixfold.¹⁹ The close link between T2DM and CVD is underscored further with recent data from the Framingham Heart Study, which showed a high lifetime risk of CVD in patients with diabetes, heightened further by obesity. During the 30-year study period, the lifetime risk of CVD in normal-weight people with diabetes was 78.6% in men and 54.8% in women; the risk increased to 86.9% in obese men with diabetes and to 78.8% in obese women with diabetes.²⁰ The NHANES data also showed that the prevalence of T2DM increased in the past decade and that patients are being diagnosed at a younger age, from a mean age of 52 years in 1988–1994 to 46 years in 1999–2000.⁷

 

 

BRIDGING THE GAP FROM PATHOPHYSIOLOGY TO UNMET NEEDS

The paradigm behind the pathophysiology of T2DM has shifted from its perception as a simple “dual-defect” disease (ie, deficiency in insulin secretion and peripheral tissue insulin resistance) to a multidimensional disorder.^1,21 This new model includes overweight/obesity, insulin resistance, qualitative and quantitative defects in insulin secretion, and dysregulation in the secretion of other hormones, including the beta-cell hormone amylin, the alpha-cell hormone glucagon, and the gastrointestinal incretin hormones GLP-1 and glucose-dependent insulinotropic polypeptide.^21–23

The major target of antidiabetes agents is glycemic control, assessed by a reduction in HbA1c, but their effects on other metabolic factors and their adverse effects differ with each agent (Table 1).³ Whereas metformin and alpha-glucosidase inhibitors may help normalize glycemia with weight-neutral effects, many other agents, including insulin and its analogues, the “glinides,” first- and second-generation sulfonylureas, and TZDs, are associated with weight gain.^23,24 In addition, the propensity to induce hypoglycemia differs among agents and clearly reflects the mechanism of action of each drug. The observed limitations of older therapies treating a progressive disease that is associated with a number of comorbid conditions supports the need for continued development of new antidiabetes agents.

CLINICAL GUIDELINES AND CV RISK FACTOR MANAGEMENT

The best strategy for managing T2DM is a comprehensive approach that addresses the fundamental core defects plus associated factors that contribute to increased CV risk. Several specialty groups have suggested guidelines and algorithms for the management of T2DM and its comorbidities. These guidelines, including the ADA standards of medical care, the AACE standards in tandem with the American College of Endocrinology guidelines, and the recent joint statement from the ADA and the European Association for the Study of Diabetes (EASD), acknowledge that the core defects of T2DM and the associated CV risk factors (eg, weight gain, obesity, hypertension, dyslipidemia) are important in developing optimal treatment strategies.^1–3 Medical nutrition guidelines advocate weight loss as a key initial step in managing T2DM and the comorbidities that lead to elevated CV risk.^25,26 The National Institutes of Health and the US Department of Health and Human Services/US Department of Agriculture advocate regular physical activity, dietary assessment, and periodic comorbidity and weight assessment for all people, not just those with T2DM or CVD.^26,27

Weight reduction

Evidence in support of effective lifestyle intervention was demonstrated in the Action for Health in Diabetes (Look AHEAD) study. After 1 year, patients with T2DM treated with intensive lifestyle intervention lost an average of 8.6% of their initial weight compared with 0.7% in patients treated only with diabetes support and education (P < 0.001). The intensive-intervention patients also had a significant drop in HbA1c (from 7.3% to 6.6%; P < 0.001) and were able to reduce their antidiabetes, antihypertensive, and lipid-lowering medications.²⁸ More recent data from the Look AHEAD study reported that overweight patients with T2DM enrolled in a weight management program experienced significant weight loss, improved physical fitness, reduced physical symptoms, and overall improvement in health-related quality of life.²⁹ Thus, weight reduction appears to be a key component in reducing CV risk and improving quality of life in most patients with T2DM.^28–30

Hypertension

Hypertension is a major risk factor for microvascular complications and CVD, and may be associated with, or be the underlying result of, nephropathy.² BP control is clearly important in reducing the morbidity and mortality associated with T2DM. The recommended BP goal in patients with T2DM is less than 130/80 mm Hg.^1,2

Hyperlipidemia

According to the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III [ATP III]), diabetes is considered a CHD risk equivalent because it confers a high risk of new CHD developing within 10 years.³¹ In addition to the NCEP–ATP III guidelines, the ADA and the AACE have set target levels for lipids in patients with diabetes, including T2DM.^1,2,31 All three organizations have defined 100 mg/dL as the target level for low-density lipoprotein.

HbA1c and lifestyle intervention

The American Heart Association and the ADA initiated a call to action for global risk assessment for CVD and diabetes.³² According to their joint scientific statement, lifestyle intervention should be reinforced at every physician visit, and HbA1c should be monitored every 3 months until it is less than 7.0% and then rechecked every 6 months. Adjustments in intervention should be made if the HbA1c level is 7.0% or higher.³ A recent joint statement from the ADA and the EASD revised an earlier treatment algorithm for the initiation of therapy in patients with T2DM; the revision includes incretin therapies (ie, GLP-1 receptor agonists) as a tier 2 option, especially in patients in whom hypoglycemia and weight gain are concerns (Figure 1).³

 

 

EVOLUTION OF ANTIDIABETES THERAPIES

Traditional antidiabetes agents used in the treatment of patients with T2DM have focused mainly on insulin secretion and insulin resistance, with treatment success defined as achieving HbA1c goals with a reduced incidence of hypoglycemia.²³ Secretagogues, such as sulfonylureas and glinides, stimulate the pancreas to release insulin. Insulin sensitizers, such as TZDs and metformin, enhance the action of insulin in muscle and fat^1,3,23 and lower hepatic glucose production. The alpha-glucosidase inhibitors alter carbohydrate absorption from the gastrointestinal tract.¹ The extent to which each agent achieves treatment success in terms of glucose lowering depends on several factors, including intrinsic attributes, duration of disease, and baseline glycemic control.³

Newer agents for the treatment of T2DM include the incretin-based therapies—GLP-1 receptor agonists and DPP-4 inhibitors—which influence mechanisms beyond increasing pancreatic insulin secretion and decreasing peripheral insulin resistance (Table 2).²² The GLP-1 signaling pathway has been leveraged by two distinct pharmacologic approaches. The first involves the use of synthetic peptides with glucoregulatory effects similar to those of endogenous GLP-1 (GLP-1 receptor agonists). The second involves the use of DPP-4 inhibitors, small molecules that inhibit the proteolytic activity of DPP-4, leading to enhanced endogenous GLP-1 concentrations.²²

GLP-1 receptor agonists

Exenatide effects. Although many agents are in development, to date exenatide is the only GLP-1 receptor agonist approved by the US Food and Drug Administration (FDA).^8,33 Exenatide is an exendin-4 GLP-1 receptor agonist with multiple glucoregulatory effects, including enhanced glucose-dependent insulin secretion, reduced glucagon secretion and food intake, and slowed gastric emptying.^22,34 Exenatide is detectable in the circulation for up to 10 hours following subcutaneous (SC) administration²² and has a greater potency in reducing plasma glucose than GLP-1 in preclinical studies.^35,36

By virtue of its beneficial effects on glycemic control, weight, BP, and lipids, exenatide addresses some of the components of the metabolic syndrome.^37–41 In pivotal 30-week studies, exenatide was associated with HbA1c reductions that ranged from –0.40% to –0.86% from baseline and decreases in body weight of approximately –1 kg to –3 kg from baseline, without severe hypoglycemia.^37–39 The percentage of patients who reached the ADA goal of HbA1c less than 7.0% at 30 weeks ranged from 24% to 34%. The addition of exenatide to TZD therapy in a 16-week study was associated with mean reductions in HbA1c of –0.98%, fasting plasma glucose (FPG) concentration of –1.69 mmol/L (–30.42 mg/dL), and body weight of –1.51 kg.⁴⁰

A posthoc analysis of an open-label extension study involving patients who completed the original 30-week placebo-controlled studies showed that 46% of patients who remained on exenatide achieved the ADA goal of HbA1c less than 7.0% at 3 years.⁴¹ Exenatide administered for up to 3.5 years was associated with sustained reductions in HbA1c of –1.0% (P < .0001) and body weight of –5.3 kg (P < .001). Pancreatic beta-cell function, assessed by homeostasis model assessment, improved, as did BP, triglyceride, high-density lipoprotein, low-density lipoprotein, and aspartate aminotransferase levels.⁴¹

Comparison with insulin analogues. Comparative studies have highlighted the contrasting effects of exenatide and insulin analogues (eg, insulin glargine and fixed-ratio insulin).^42–45 In a 26-week trial comparing exenatide with insulin glargine in subjects with T2DM, both agents resulted in similar decreases in HbA1c. Exenatide was also associated with a –2.3-kg weight reduction, whereas insulin glargine was associated with a +1.8-kg weight gain.⁴² Although rates of symptomatic hypoglycemia were similar, there were fewer cases of nocturnal hypoglycemia with exenatide (0.9 event/patient-year vs 2.4 events/patient-year with insulin).

In a 32-week study comparing exenatide BID with titrated insulin glargine QD, the HbA1c reductions for exenatide and insulin glargine were comparable. However, body weight decreased –4.2 kg over two 16-week treatment periods with exenatide, but increased +3.3 kg over the same periods with the basal insulin analogue.⁴³ The incidence of hypoglycemia was lower with exenatide than with insulin glargine (14.7% vs 25.2%), although the difference was not statistically significant.

In another study that compared exenatide with biphasic insulin aspart, patients who were treated with exenatide also lost weight while those who received the fast-acting insulin analogue gained weight (between-group difference, –5.4 kg). Patients treated with exenatide also demonstrated greater reductions in postprandial plasma glucose (PPG) excursions following their morning (P < .001), midday (P = .002), and evening meals (P < .001).⁴⁴ Overall, hypoglycemia rates were similar at study end between exenatide and insulin aspart (4.7 events/patient-year vs 5.6 events/patient-year). In all of these studies, significant gastrointestinal adverse events (nausea and vomiting) occurred more frequently with exenatide, and more patients withdrew from exenatide than from insulin.

Formulations in development. Other advances in GLP-1 receptor agonist therapy include novel formulations under clinical development, such as exenatide once weekly^36,46 and liraglutide, a human analogue GLP-1 receptor agonist formulated for once-daily administration.^47,48 In a 52-week study in patients with T2DM, liraglutide significantly reduced HbA1c; the 1.2-mg SC QD dosage reduced HBA1c by –0.84% (P = .0014) and the 1.8-mg SC QD dosage by –1.14% (P < .0001). In comparison, glimepiride 8 mg orally QD achieved a –0.51% reduction. Liraglutide was also associated with greater reductions in weight, hypoglycemia, and systolic BP than glimepiride.⁴⁷

A 26-week study compared liraglutide (0.6, 1.2, and 1.8 mg SC QD), placebo, and glimepiride 4 mg QD in combination with metformin 1 g BID. HbA1c was reduced significantly in all liraglutide groups compared with placebo (P < .0001). Mean HbA1c decreased –1.0% with liraglutide 1.2 mg and 1.8 mg and with glimepiride; it decreased –0.7% with liraglutide 0.6 mg; and it increased +0.1% with placebo. Body weight decreased –1.8 kg to –2.8 kg in all liraglutide groups but increased +1.0 kg in the glimepiride group (P < .0001). The incidence of minor hypoglycemia with liraglutide (~3%) was comparable to that observed with placebo but less than that with glimepiride (17%; P < .001).⁴⁸

A once-weekly long-acting release (LAR) formulation of exenatide submitted to the FDA for approval may provide enhanced glycemic and weight control, potentially improving patient acceptance and adherence.^36,46 In a 15-week study, exenatide once weekly produced significant reductions in HbA1c, FPG, PPG, and body weight. There were no withdrawals due to adverse events, and the formation of anti-exenatide antibodies was not predictive of therapeutic end point response or adverse safety outcome. Instances of hypoglycemia were mild and not dose related.³⁶ In a 30-week study comparing exenatide LAR once weekly with exenatide BID, patients given exenatide LAR once weekly had significantly greater HbA1c reductions than did patients given exenatide BID (–1.9% vs –1.5%; P = .0023). Treatment adherence was 98% with both exenatide regimens, and no episodes of major hypoglycemia occurred with either formulation regardless of background sulfonylurea use. Favorable effects on BP and lipid profile were observed with both exenatide regimens.⁴⁶

 

 

DPP-4 inhibitors

The DPP-4 inhibitors (commonly called gliptins) inhibit the proteolytic cleavage of circulating GLP-1 by binding to the DPP-4 enzyme, increasing the concentration of endogenous GLP-1 approximately two- to threefold.^49–51 These concentrations result in more prompt and appropriate secretion of insulin and suppression of glucagon in response to a carbohydrate-containing snack or meal, with the change in glucagon correlating linearly with improved glucose tolerance.⁵¹

DPP-4 inhibitors, which are given orally, include sitagliptin and saxagliptin (approved in the United States) and vildagliptin (not approved in the United States but used in the European Union and Latin America).^8,22,33,52 Sitagliptin can be used either as monotherapy or in combination with metformin or a TZD.^8,49–55 Recently, a single-tablet formulation of sitagliptin plus metformin was granted regulatory approval.⁸

When used alone or in combination with metformin or pioglitazone, sitagliptin has been associated with significant reductions in HbA1c (of ~0.5% to 0.6% when used alone, ~0.7% with metformin, and ~0.9% with pioglitazone [P < .001 vs placebo]), with hypoglycemia occurring in 1.3% or less of the population.⁵⁴ In an 18-week study in which patients with T2DM who were inadequately controlled with metformin monotherapy were randomized to receive add-on sitagliptin (100 mg QD), rosiglitazone (8 mg QD), or placebo, sitagliptin reduced HbA1c –0.73% (P < .001 vs placebo) and reduced body weight –0.4 kg, while rosiglitazone reduced HbA1c –0.79% and increased body weight +1.5 kg.⁵⁵

To evaluate the effectiveness of sitagliptin and metformin as initial therapy, a 54-week study was completed in 885 patients with T2DM and inadequate glycemic control (HbA1c 7.5–11%) on diet and exercise.⁵⁶ Patients were evaluated on monotherapy with either sitagliptin (100 mg QD) or metformin (1 g or 2 g QD), or on initial therapy with the two in combination (sitagliptin 100 mg + metformin 1 mg or 2 mg QD). At week 54, in the all-patients-treated analysis, mean changes in HbA1c from baseline were –1.8% with sitagliptin plus metformin 2 g QD, –1.4% with sitagliptin plus metformin 1 g QD, –1.3% with metformin 2 g QD monotherapy, –1.0% with metformin 1 g QD monotherapy, and –0.8% with sitagliptin 100 mg QD monotherapy.

All treatments improved measures of beta-cell function (eg, homeostasis model assessment [HOMA]-beta, proinsulin/insulin ratio). Mean body weight decreased from baseline in the combination and metformin monotherapy groups and was unchanged from baseline in the sitagliptin monotherapy group. The incidence of hypoglycemia was low (1%–3%) across treatment groups. The incidence of gastrointestinal adverse experiences was evaluated with the coadministration of sitagliptin and metformin and appeared similar to that observed with use of metformin as monotherapy.⁵⁶ Thus, this study suggested that an initial combination of a DPP-IV inhibitor with metformin can improve glycemic control and markers of beta-cell function in patients with T2DM.

Incretin-based therapies compared

Studies in both healthy individuals and in patients with T2DM have shown that oral DPP-4 inhibitors such as sitagliptin increase endogenous GLP-1 concentrations by about twofold compared with placebo.^22,50 The pharma­cologic concentration of subcutaneously administered exenatide available for activating the GLP-1 receptor is significantly greater than the increased endogenous GLP-1 concentrations achieved with sitagliptin. In a recent clinical study comparing exenatide and sitagliptin in patients with T2DM, the mean 2-hour plasma concentration for exenatide was 64 pM compared with the mean 2-hour postprandial GLP-1 concentration of 15 pM for sitagliptin (baseline GLP-1 concentration was 7.2 pM).⁵⁷ While both agents were shown to be effective, exenatide appeared to have had a greater effect than sitagliptin in increasing insulin secretion and reducing postprandial glucagon secretion, leading to significantly (P < 0.0001) greater reductions in PPG.⁵⁷

Sitagliptin has been minimally associated with nausea, whereas patients who take exenatide need to be informed of the risk of usually mild to moderate, but sometimes severe, nausea and vomiting that tends to decrease over time.

For a detailed comparison of the effects of GLP-1 receptor agonists and DPP-4 inhibitors on HbA1c, weight, and hypoglycemia, see “Advances in therapy for type 2 diabetes: GLP–1 receptor agonists and DPP–4 inhibitors.”

CONCLUSION

Despite advances in diagnosis and treatment, T2DM, overweight/obesity, CVD, and their complications remain major public health burdens worldwide. The concepts that explain the pathophysiology of T2DM include the contribution of various factors beyond insulin secretion and insulin resistance, such as the role of incretin hormones in disease progression. A comprehensive approach to managing patients with T2DM requires targeting the fundamental defects of the disease and its comorbidities. Newer agents, including incretin-based therapies such as GLP-1 receptor agonists and DPP-4 inhibitors, address the fundamental defects of T2DM. The definition of treatment success in the management of T2DM will be redefined as more data become available on agents that exert beneficial effects not only on glycemia but on parameters that may influence overall CV health, such as weight, BP, and lipid profiles.

According to the American Association of Clinical Endocrinologists (AACE) and the American Diabetes Association (ADA), glycosylated hemoglobin (HbA1c) in patients with diabetes should be maintained at 6.5% or less (AACE) or at less than 7.0% (ADA). Both organizations support an aggressive stepwise approach that includes medication and lifestyle modification, with strategies and clinical attention devoted to avoiding significant hypoglycemia.^1,2 Yet, despite the introduction of new antidiabetes agents, most current management strategies are offset by limitations in achieving and maintaining glycemic targets needed to provide optimal care for patients with diabetes, more than 90% of whom have type 2 diabetes mellitus (T2DM).^3,4

Nationally, glycemic control among patients with T2DM has improved but is still far from optimal. According to data from the 1999–2000 National Health and Nutrition Examination Survey (NHANES), glycemic control (HbA1c < 7.0%) rates were 35.8% for patients with T2DM.⁵ In a more recent report (NHANES 1999–2004), fewer than half (48.4%) of adult patients with diagnosed diabetes achieved HbA1c levels below 7.0%.^5,6 Factors contributing to these data include earlier onset and earlier detection of T2DM.⁷

CHANGING TREATMENT TRENDS

Available treatments for patients with T2DM include secretagogues, such as sulfonylureas and “glinides” (repaglinide and nateglinide), metformin, thiazolidinediones (TZDs), and dipeptidyl peptidase–4 (DPP-4) inhibitors among oral medications, and insulin and glucagon-like peptide–1 (GLP-1) receptor agonists among parenterally administered agents. According to the latest published data on prescribing patterns for patients with T2DM, analyses of the National Disease and Therapeutic Index (1994–2007) and the National Prescription Audit (2001–2007), sulfonylurea use decreased from 67% of treatment visits in 1994 to 34% of visits in 2007.⁸ By 2007, metformin, used in 54% of treatment visits, and TZDs, used in 28%, were the most frequently administered antidiabetes agents. Insulin use declined from 38% of visits during which a treatment was administered in 1994 to 25% of visits in 2000, but had increased subsequently to 28% of visits in 2007.

SIGNIFICANCE OF CARDIOVASCULAR RISK

Clinical research has suggested that focusing solely on improving glycemic control may be insufficient to reduce overall morbidity and mortality associated with diabetes. Specifically, data from recent studies, including the Action to Control Cardiovascular Risk in Diabetes (ACCORD), the Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE), and the Veterans Affairs Diabetes Trial (VADT), emphasized that lowering HbA1c below 7% in a high-risk population of individuals with T2DM did not improve cardiovascular (CV) outcomes.^9–11 The observations confirm that risk factors, including weight, blood pressure (BP), and lipid levels, are vitally important in reducing morbidity and mortality in this population. This perception is further underscored by the NHANES 1999–2004 data, which showed poor concurrent control of HbA1c, BP, and lipids; only 13.2% of patients with diagnosed diabetes achieved all three target goals simultaneously.⁶ Similarly, a nationwide survey in Norway showed that only 13% of patients with T2DM concurrently achieved goals for HbA1c, BP, and lipids.¹²

In the Danish Steno-2 Study, patients with T2DM and persistent microalbuminuria were treated with either intensive target-driven therapy using multiple drugs or conventional multifactorial treatment. Over a mean period of 13.3 years (7.8 years of treatment plus 5.5 years of follow-up), intensive multifactorial intervention to control multiple CV risk factors, including HbA1c, BP, and lipids, was associated with a lower risk of death from CV causes (hazard ratio [HR], 0.43; 95% confidence interval [CI], 0.19 to 0.94; P = .04) and a lower risk of CV events (HR, 0.41; 95% CI, 0.25 to 0.67; P < .001) than was conventional therapy.¹³

This article clarifies the redefinition of treatment success in patients with T2DM based on targeting the underlying physiologic defects of the disease.

T2DM, OVERWEIGHT/OBESITY, AND CV DISEASE: CLOSELY LINKED

The incidence and prevalence of T2DM, overweight/obesity, and CV disease (CVD) are increasing worldwide. It is estimated that the worldwide prevalence of diabetes will increase from 171 million in 2000 to 366 million by 2030¹⁴; T2DM increases the risk of morbidity and mortality from microvascular (eg, neuropathic, retinopathic, nephropathic) and macrovascular (eg, coronary, peripheral vascular disease) complications.¹⁵ According to a Michigan health maintenance organization study (N = 1,364), the median annual direct cost of medical care for Caucasian patients with T2DM who were diet controlled, had a body mass index (BMI) of 30 kg/m² or higher, and had no vascular complications was estimated to be $1,700 for men and $2,100 for women.¹⁶ The actual cost of care for patients with T2DM may be much higher, since most patients present with multiple CV risk factors in addition to being overweight.

NHANES data show that approximately two-thirds of Americans are either overweight or obese¹⁷; overweight/obesity affects about 80% of adults diagnosed with T2DM.¹⁸ Overweight or obesity can increase the risk for developing T2DM by more than 90-fold and, in women, it can increase the risk for developing coronary heart disease (CHD) by sixfold.¹⁹ The close link between T2DM and CVD is underscored further with recent data from the Framingham Heart Study, which showed a high lifetime risk of CVD in patients with diabetes, heightened further by obesity. During the 30-year study period, the lifetime risk of CVD in normal-weight people with diabetes was 78.6% in men and 54.8% in women; the risk increased to 86.9% in obese men with diabetes and to 78.8% in obese women with diabetes.²⁰ The NHANES data also showed that the prevalence of T2DM increased in the past decade and that patients are being diagnosed at a younger age, from a mean age of 52 years in 1988–1994 to 46 years in 1999–2000.⁷

 

 

BRIDGING THE GAP FROM PATHOPHYSIOLOGY TO UNMET NEEDS

The paradigm behind the pathophysiology of T2DM has shifted from its perception as a simple “dual-defect” disease (ie, deficiency in insulin secretion and peripheral tissue insulin resistance) to a multidimensional disorder.^1,21 This new model includes overweight/obesity, insulin resistance, qualitative and quantitative defects in insulin secretion, and dysregulation in the secretion of other hormones, including the beta-cell hormone amylin, the alpha-cell hormone glucagon, and the gastrointestinal incretin hormones GLP-1 and glucose-dependent insulinotropic polypeptide.^21–23

The major target of antidiabetes agents is glycemic control, assessed by a reduction in HbA1c, but their effects on other metabolic factors and their adverse effects differ with each agent (Table 1).³ Whereas metformin and alpha-glucosidase inhibitors may help normalize glycemia with weight-neutral effects, many other agents, including insulin and its analogues, the “glinides,” first- and second-generation sulfonylureas, and TZDs, are associated with weight gain.^23,24 In addition, the propensity to induce hypoglycemia differs among agents and clearly reflects the mechanism of action of each drug. The observed limitations of older therapies treating a progressive disease that is associated with a number of comorbid conditions supports the need for continued development of new antidiabetes agents.

CLINICAL GUIDELINES AND CV RISK FACTOR MANAGEMENT

The best strategy for managing T2DM is a comprehensive approach that addresses the fundamental core defects plus associated factors that contribute to increased CV risk. Several specialty groups have suggested guidelines and algorithms for the management of T2DM and its comorbidities. These guidelines, including the ADA standards of medical care, the AACE standards in tandem with the American College of Endocrinology guidelines, and the recent joint statement from the ADA and the European Association for the Study of Diabetes (EASD), acknowledge that the core defects of T2DM and the associated CV risk factors (eg, weight gain, obesity, hypertension, dyslipidemia) are important in developing optimal treatment strategies.^1–3 Medical nutrition guidelines advocate weight loss as a key initial step in managing T2DM and the comorbidities that lead to elevated CV risk.^25,26 The National Institutes of Health and the US Department of Health and Human Services/US Department of Agriculture advocate regular physical activity, dietary assessment, and periodic comorbidity and weight assessment for all people, not just those with T2DM or CVD.^26,27

Weight reduction

Evidence in support of effective lifestyle intervention was demonstrated in the Action for Health in Diabetes (Look AHEAD) study. After 1 year, patients with T2DM treated with intensive lifestyle intervention lost an average of 8.6% of their initial weight compared with 0.7% in patients treated only with diabetes support and education (P < 0.001). The intensive-intervention patients also had a significant drop in HbA1c (from 7.3% to 6.6%; P < 0.001) and were able to reduce their antidiabetes, antihypertensive, and lipid-lowering medications.²⁸ More recent data from the Look AHEAD study reported that overweight patients with T2DM enrolled in a weight management program experienced significant weight loss, improved physical fitness, reduced physical symptoms, and overall improvement in health-related quality of life.²⁹ Thus, weight reduction appears to be a key component in reducing CV risk and improving quality of life in most patients with T2DM.^28–30

Hypertension

Hypertension is a major risk factor for microvascular complications and CVD, and may be associated with, or be the underlying result of, nephropathy.² BP control is clearly important in reducing the morbidity and mortality associated with T2DM. The recommended BP goal in patients with T2DM is less than 130/80 mm Hg.^1,2

Hyperlipidemia

According to the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III [ATP III]), diabetes is considered a CHD risk equivalent because it confers a high risk of new CHD developing within 10 years.³¹ In addition to the NCEP–ATP III guidelines, the ADA and the AACE have set target levels for lipids in patients with diabetes, including T2DM.^1,2,31 All three organizations have defined 100 mg/dL as the target level for low-density lipoprotein.

HbA1c and lifestyle intervention

The American Heart Association and the ADA initiated a call to action for global risk assessment for CVD and diabetes.³² According to their joint scientific statement, lifestyle intervention should be reinforced at every physician visit, and HbA1c should be monitored every 3 months until it is less than 7.0% and then rechecked every 6 months. Adjustments in intervention should be made if the HbA1c level is 7.0% or higher.³ A recent joint statement from the ADA and the EASD revised an earlier treatment algorithm for the initiation of therapy in patients with T2DM; the revision includes incretin therapies (ie, GLP-1 receptor agonists) as a tier 2 option, especially in patients in whom hypoglycemia and weight gain are concerns (Figure 1).³

 

 

EVOLUTION OF ANTIDIABETES THERAPIES

Traditional antidiabetes agents used in the treatment of patients with T2DM have focused mainly on insulin secretion and insulin resistance, with treatment success defined as achieving HbA1c goals with a reduced incidence of hypoglycemia.²³ Secretagogues, such as sulfonylureas and glinides, stimulate the pancreas to release insulin. Insulin sensitizers, such as TZDs and metformin, enhance the action of insulin in muscle and fat^1,3,23 and lower hepatic glucose production. The alpha-glucosidase inhibitors alter carbohydrate absorption from the gastrointestinal tract.¹ The extent to which each agent achieves treatment success in terms of glucose lowering depends on several factors, including intrinsic attributes, duration of disease, and baseline glycemic control.³

Newer agents for the treatment of T2DM include the incretin-based therapies—GLP-1 receptor agonists and DPP-4 inhibitors—which influence mechanisms beyond increasing pancreatic insulin secretion and decreasing peripheral insulin resistance (Table 2).²² The GLP-1 signaling pathway has been leveraged by two distinct pharmacologic approaches. The first involves the use of synthetic peptides with glucoregulatory effects similar to those of endogenous GLP-1 (GLP-1 receptor agonists). The second involves the use of DPP-4 inhibitors, small molecules that inhibit the proteolytic activity of DPP-4, leading to enhanced endogenous GLP-1 concentrations.²²

GLP-1 receptor agonists

Exenatide effects. Although many agents are in development, to date exenatide is the only GLP-1 receptor agonist approved by the US Food and Drug Administration (FDA).^8,33 Exenatide is an exendin-4 GLP-1 receptor agonist with multiple glucoregulatory effects, including enhanced glucose-dependent insulin secretion, reduced glucagon secretion and food intake, and slowed gastric emptying.^22,34 Exenatide is detectable in the circulation for up to 10 hours following subcutaneous (SC) administration²² and has a greater potency in reducing plasma glucose than GLP-1 in preclinical studies.^35,36

By virtue of its beneficial effects on glycemic control, weight, BP, and lipids, exenatide addresses some of the components of the metabolic syndrome.^37–41 In pivotal 30-week studies, exenatide was associated with HbA1c reductions that ranged from –0.40% to –0.86% from baseline and decreases in body weight of approximately –1 kg to –3 kg from baseline, without severe hypoglycemia.^37–39 The percentage of patients who reached the ADA goal of HbA1c less than 7.0% at 30 weeks ranged from 24% to 34%. The addition of exenatide to TZD therapy in a 16-week study was associated with mean reductions in HbA1c of –0.98%, fasting plasma glucose (FPG) concentration of –1.69 mmol/L (–30.42 mg/dL), and body weight of –1.51 kg.⁴⁰

A posthoc analysis of an open-label extension study involving patients who completed the original 30-week placebo-controlled studies showed that 46% of patients who remained on exenatide achieved the ADA goal of HbA1c less than 7.0% at 3 years.⁴¹ Exenatide administered for up to 3.5 years was associated with sustained reductions in HbA1c of –1.0% (P < .0001) and body weight of –5.3 kg (P < .001). Pancreatic beta-cell function, assessed by homeostasis model assessment, improved, as did BP, triglyceride, high-density lipoprotein, low-density lipoprotein, and aspartate aminotransferase levels.⁴¹

Comparison with insulin analogues. Comparative studies have highlighted the contrasting effects of exenatide and insulin analogues (eg, insulin glargine and fixed-ratio insulin).^42–45 In a 26-week trial comparing exenatide with insulin glargine in subjects with T2DM, both agents resulted in similar decreases in HbA1c. Exenatide was also associated with a –2.3-kg weight reduction, whereas insulin glargine was associated with a +1.8-kg weight gain.⁴² Although rates of symptomatic hypoglycemia were similar, there were fewer cases of nocturnal hypoglycemia with exenatide (0.9 event/patient-year vs 2.4 events/patient-year with insulin).

In a 32-week study comparing exenatide BID with titrated insulin glargine QD, the HbA1c reductions for exenatide and insulin glargine were comparable. However, body weight decreased –4.2 kg over two 16-week treatment periods with exenatide, but increased +3.3 kg over the same periods with the basal insulin analogue.⁴³ The incidence of hypoglycemia was lower with exenatide than with insulin glargine (14.7% vs 25.2%), although the difference was not statistically significant.

In another study that compared exenatide with biphasic insulin aspart, patients who were treated with exenatide also lost weight while those who received the fast-acting insulin analogue gained weight (between-group difference, –5.4 kg). Patients treated with exenatide also demonstrated greater reductions in postprandial plasma glucose (PPG) excursions following their morning (P < .001), midday (P = .002), and evening meals (P < .001).⁴⁴ Overall, hypoglycemia rates were similar at study end between exenatide and insulin aspart (4.7 events/patient-year vs 5.6 events/patient-year). In all of these studies, significant gastrointestinal adverse events (nausea and vomiting) occurred more frequently with exenatide, and more patients withdrew from exenatide than from insulin.

Formulations in development. Other advances in GLP-1 receptor agonist therapy include novel formulations under clinical development, such as exenatide once weekly^36,46 and liraglutide, a human analogue GLP-1 receptor agonist formulated for once-daily administration.^47,48 In a 52-week study in patients with T2DM, liraglutide significantly reduced HbA1c; the 1.2-mg SC QD dosage reduced HBA1c by –0.84% (P = .0014) and the 1.8-mg SC QD dosage by –1.14% (P < .0001). In comparison, glimepiride 8 mg orally QD achieved a –0.51% reduction. Liraglutide was also associated with greater reductions in weight, hypoglycemia, and systolic BP than glimepiride.⁴⁷

A 26-week study compared liraglutide (0.6, 1.2, and 1.8 mg SC QD), placebo, and glimepiride 4 mg QD in combination with metformin 1 g BID. HbA1c was reduced significantly in all liraglutide groups compared with placebo (P < .0001). Mean HbA1c decreased –1.0% with liraglutide 1.2 mg and 1.8 mg and with glimepiride; it decreased –0.7% with liraglutide 0.6 mg; and it increased +0.1% with placebo. Body weight decreased –1.8 kg to –2.8 kg in all liraglutide groups but increased +1.0 kg in the glimepiride group (P < .0001). The incidence of minor hypoglycemia with liraglutide (~3%) was comparable to that observed with placebo but less than that with glimepiride (17%; P < .001).⁴⁸

A once-weekly long-acting release (LAR) formulation of exenatide submitted to the FDA for approval may provide enhanced glycemic and weight control, potentially improving patient acceptance and adherence.^36,46 In a 15-week study, exenatide once weekly produced significant reductions in HbA1c, FPG, PPG, and body weight. There were no withdrawals due to adverse events, and the formation of anti-exenatide antibodies was not predictive of therapeutic end point response or adverse safety outcome. Instances of hypoglycemia were mild and not dose related.³⁶ In a 30-week study comparing exenatide LAR once weekly with exenatide BID, patients given exenatide LAR once weekly had significantly greater HbA1c reductions than did patients given exenatide BID (–1.9% vs –1.5%; P = .0023). Treatment adherence was 98% with both exenatide regimens, and no episodes of major hypoglycemia occurred with either formulation regardless of background sulfonylurea use. Favorable effects on BP and lipid profile were observed with both exenatide regimens.⁴⁶

 

 

DPP-4 inhibitors

The DPP-4 inhibitors (commonly called gliptins) inhibit the proteolytic cleavage of circulating GLP-1 by binding to the DPP-4 enzyme, increasing the concentration of endogenous GLP-1 approximately two- to threefold.^49–51 These concentrations result in more prompt and appropriate secretion of insulin and suppression of glucagon in response to a carbohydrate-containing snack or meal, with the change in glucagon correlating linearly with improved glucose tolerance.⁵¹

DPP-4 inhibitors, which are given orally, include sitagliptin and saxagliptin (approved in the United States) and vildagliptin (not approved in the United States but used in the European Union and Latin America).^8,22,33,52 Sitagliptin can be used either as monotherapy or in combination with metformin or a TZD.^8,49–55 Recently, a single-tablet formulation of sitagliptin plus metformin was granted regulatory approval.⁸

When used alone or in combination with metformin or pioglitazone, sitagliptin has been associated with significant reductions in HbA1c (of ~0.5% to 0.6% when used alone, ~0.7% with metformin, and ~0.9% with pioglitazone [P < .001 vs placebo]), with hypoglycemia occurring in 1.3% or less of the population.⁵⁴ In an 18-week study in which patients with T2DM who were inadequately controlled with metformin monotherapy were randomized to receive add-on sitagliptin (100 mg QD), rosiglitazone (8 mg QD), or placebo, sitagliptin reduced HbA1c –0.73% (P < .001 vs placebo) and reduced body weight –0.4 kg, while rosiglitazone reduced HbA1c –0.79% and increased body weight +1.5 kg.⁵⁵

To evaluate the effectiveness of sitagliptin and metformin as initial therapy, a 54-week study was completed in 885 patients with T2DM and inadequate glycemic control (HbA1c 7.5–11%) on diet and exercise.⁵⁶ Patients were evaluated on monotherapy with either sitagliptin (100 mg QD) or metformin (1 g or 2 g QD), or on initial therapy with the two in combination (sitagliptin 100 mg + metformin 1 mg or 2 mg QD). At week 54, in the all-patients-treated analysis, mean changes in HbA1c from baseline were –1.8% with sitagliptin plus metformin 2 g QD, –1.4% with sitagliptin plus metformin 1 g QD, –1.3% with metformin 2 g QD monotherapy, –1.0% with metformin 1 g QD monotherapy, and –0.8% with sitagliptin 100 mg QD monotherapy.

All treatments improved measures of beta-cell function (eg, homeostasis model assessment [HOMA]-beta, proinsulin/insulin ratio). Mean body weight decreased from baseline in the combination and metformin monotherapy groups and was unchanged from baseline in the sitagliptin monotherapy group. The incidence of hypoglycemia was low (1%–3%) across treatment groups. The incidence of gastrointestinal adverse experiences was evaluated with the coadministration of sitagliptin and metformin and appeared similar to that observed with use of metformin as monotherapy.⁵⁶ Thus, this study suggested that an initial combination of a DPP-IV inhibitor with metformin can improve glycemic control and markers of beta-cell function in patients with T2DM.

Incretin-based therapies compared

Studies in both healthy individuals and in patients with T2DM have shown that oral DPP-4 inhibitors such as sitagliptin increase endogenous GLP-1 concentrations by about twofold compared with placebo.^22,50 The pharma­cologic concentration of subcutaneously administered exenatide available for activating the GLP-1 receptor is significantly greater than the increased endogenous GLP-1 concentrations achieved with sitagliptin. In a recent clinical study comparing exenatide and sitagliptin in patients with T2DM, the mean 2-hour plasma concentration for exenatide was 64 pM compared with the mean 2-hour postprandial GLP-1 concentration of 15 pM for sitagliptin (baseline GLP-1 concentration was 7.2 pM).⁵⁷ While both agents were shown to be effective, exenatide appeared to have had a greater effect than sitagliptin in increasing insulin secretion and reducing postprandial glucagon secretion, leading to significantly (P < 0.0001) greater reductions in PPG.⁵⁷

Sitagliptin has been minimally associated with nausea, whereas patients who take exenatide need to be informed of the risk of usually mild to moderate, but sometimes severe, nausea and vomiting that tends to decrease over time.

For a detailed comparison of the effects of GLP-1 receptor agonists and DPP-4 inhibitors on HbA1c, weight, and hypoglycemia, see “Advances in therapy for type 2 diabetes: GLP–1 receptor agonists and DPP–4 inhibitors.”

CONCLUSION

Despite advances in diagnosis and treatment, T2DM, overweight/obesity, CVD, and their complications remain major public health burdens worldwide. The concepts that explain the pathophysiology of T2DM include the contribution of various factors beyond insulin secretion and insulin resistance, such as the role of incretin hormones in disease progression. A comprehensive approach to managing patients with T2DM requires targeting the fundamental defects of the disease and its comorbidities. Newer agents, including incretin-based therapies such as GLP-1 receptor agonists and DPP-4 inhibitors, address the fundamental defects of T2DM. The definition of treatment success in the management of T2DM will be redefined as more data become available on agents that exert beneficial effects not only on glycemia but on parameters that may influence overall CV health, such as weight, BP, and lipid profiles.

As hospitalists take on more demanding leadership roles, the climb up the career ladder evolves into a juggling act: Hospitalists typically try to handle a full patient load as well as new administrative duties.

If a hospitalist continues to ascend, those administrative duties can begin to consume the schedule. The individual—and the group—could face important decisions about priorities, schedules, and money.

“Hospital medicine is only ten years old; we’re still trying to figure this out,” says Mary Jo Gorman, MD, MBA, chief executive officer of Advanced ICU Care in St. Louis and a past president of SHM. “It’s always a challenge. You identify that you have a need for someone to take charge of an administrative task, but it can take as long as a year to free up [the hospitalist’s] time so that it can get done.”

If you have found yourself in this position, you know that something has to give. “I’ve seen high-energy physicians who think that they can do it all—and they had to,” says Joan C. Faro, MD, FACP, MBA, chief medical officer at John T. Mather Memorial Hospital in Port Jefferson, N.Y. “That is not sustainable. It can’t last forever.”

The question is, how can a hospitalist effectively balance their clinical and administrative duties? Furthermore, what happens when the scales tip in favor—and to the detriment—of one or the other?

When the Juggling Begins

Hospitalists usually add “extra” duties to their normal workloads to advance their careers. Few relinquish their clinical duties as they join committees, further their training, lead a research project, or take on administrative duties.

Dr. Faro says a hospitalist should be able to “head up a focused project or serve on committees” and still be able to meet all their clinical duties. “Once you get beyond that, you need a certain amount of protected time” for administrative or project work, she says. “And when you start to have people reporting to you, you absolutely need that protected time.”

Assigning administrative tasks to physicians who regularly see patients depends on the group structure and requires a clearly defined job description. “If a group is really going to make this work, then you have to pay people for that extra time,” Dr. Gorman says.

Ideally, HM groups have job descriptions for physicians who are called upon to see patients and handle administrative duties. Contracts should include specifications for “protected time,” as well as compensation for new responsibilities.

Clinical-Hour Cutbacks

As administrative duties grow, something has to give. Hospitalists who want to pursue positions of leadership know that that something is hours spent delivering patient care. “If you’re a hospitalist-administrator who wants to make the leap to vice president or department chair or chief medical officer, you need to devote a lot of time to your administrative work,” Dr. Faro says. “You can’t make that leap without putting in those hours.”

So what is a reasonable division of time for, say, the director of an HM program or department? “It’s impossible to pinpoint, but I’d say roughly that [a director] should spend not less than 25% or 30% of their time, and certainly not more than 50% of their time, on clinical work,” Dr. Faro estimates.

 

Even upper-level physician-administrators should maintain a clinical practice simply to monitor the work their department is doing. “It’s not about [clinical] skills as much as it is about whether you can relate to physicians’ day-to-day work, to their frustrations,” Dr. Gorman says. “That’s a management challenge no matter who you are. For example, if hospitalists are complaining about a new EMR [electronic medical record] system, are you going to say, ‘Oh, just put up with it; it’s not that bad. It will be fine’? Or are you out there trying it and saying, ‘Holy cow, this is really inefficient. We have to change this’?”

On the flip side, how much time should be devoted to administrative tasks? The answer depends on the size of your program and the amount of work you have to do, Dr. Faro says. Group directors and department heads normally make themselves available during regular weekday hours. That usually means you’ll have to fit in your clinical work around meetings, budgets, and presentations.

Can You Give Up Clinical Duties?

It’s natural for physicians to be reluctant to relinquish patient care; some reach a point where they have to make the tough decision to stop clinical work altogether.

“You may figure out that you want to pursue an administrative role, but you don’t want to give up clinical work,” says Dr. Gorman, who spent 15 years juggling a full clinical schedule with administrative duties before she became a full-time administrator. “You get plenty of opportunities to make that decision as you’re crossing back and forth.”

You might want to evaluate your options and make the choice sooner rather than later. Once you’re in administration, the decision might be forced upon you. “Eventually, you’ll find that critical things are happening all hours of the day, any given day of the week, in administration as well as clinical practice,” Dr. Faro says. “There’s a point at which you realize that part-time [administrative work] just doesn’t work. You realize that your expertise and guidance are needed.”

Dr. Gorman warns that there are risks and changes involved with becoming a full-time administrator. Once you decide to give up your clinical practice and go the leadership route, your career is “in the hands of someone else,” she points out. “Your position could be eliminated. You could be fired or replaced. … That is a concern. A lot of people keep their hand in on clinical skills for that reason.” You also might find that advancing a management career requires moving to a new organization or a different part of the country.

On the other hand, the rewards of a career in administration can’t be overlooked. “It’s very satisfying personally,” Dr. Faro says. “It’s inventive; you’re constantly solving problems that didn’t exist yesterday.

 

“It’s a different kind of job satisfaction. It’s a very personal decision. There are people who realize that this just isn’t for them.” TH

Jane Jerrard is a freelance writer based in Chicago.

As hospitalists take on more demanding leadership roles, the climb up the career ladder evolves into a juggling act: Hospitalists typically try to handle a full patient load as well as new administrative duties.

If a hospitalist continues to ascend, those administrative duties can begin to consume the schedule. The individual—and the group—could face important decisions about priorities, schedules, and money.

“Hospital medicine is only ten years old; we’re still trying to figure this out,” says Mary Jo Gorman, MD, MBA, chief executive officer of Advanced ICU Care in St. Louis and a past president of SHM. “It’s always a challenge. You identify that you have a need for someone to take charge of an administrative task, but it can take as long as a year to free up [the hospitalist’s] time so that it can get done.”

If you have found yourself in this position, you know that something has to give. “I’ve seen high-energy physicians who think that they can do it all—and they had to,” says Joan C. Faro, MD, FACP, MBA, chief medical officer at John T. Mather Memorial Hospital in Port Jefferson, N.Y. “That is not sustainable. It can’t last forever.”

The question is, how can a hospitalist effectively balance their clinical and administrative duties? Furthermore, what happens when the scales tip in favor—and to the detriment—of one or the other?

When the Juggling Begins

Hospitalists usually add “extra” duties to their normal workloads to advance their careers. Few relinquish their clinical duties as they join committees, further their training, lead a research project, or take on administrative duties.

Dr. Faro says a hospitalist should be able to “head up a focused project or serve on committees” and still be able to meet all their clinical duties. “Once you get beyond that, you need a certain amount of protected time” for administrative or project work, she says. “And when you start to have people reporting to you, you absolutely need that protected time.”

Assigning administrative tasks to physicians who regularly see patients depends on the group structure and requires a clearly defined job description. “If a group is really going to make this work, then you have to pay people for that extra time,” Dr. Gorman says.

Ideally, HM groups have job descriptions for physicians who are called upon to see patients and handle administrative duties. Contracts should include specifications for “protected time,” as well as compensation for new responsibilities.

Clinical-Hour Cutbacks

As administrative duties grow, something has to give. Hospitalists who want to pursue positions of leadership know that that something is hours spent delivering patient care. “If you’re a hospitalist-administrator who wants to make the leap to vice president or department chair or chief medical officer, you need to devote a lot of time to your administrative work,” Dr. Faro says. “You can’t make that leap without putting in those hours.”

So what is a reasonable division of time for, say, the director of an HM program or department? “It’s impossible to pinpoint, but I’d say roughly that [a director] should spend not less than 25% or 30% of their time, and certainly not more than 50% of their time, on clinical work,” Dr. Faro estimates.

 

Even upper-level physician-administrators should maintain a clinical practice simply to monitor the work their department is doing. “It’s not about [clinical] skills as much as it is about whether you can relate to physicians’ day-to-day work, to their frustrations,” Dr. Gorman says. “That’s a management challenge no matter who you are. For example, if hospitalists are complaining about a new EMR [electronic medical record] system, are you going to say, ‘Oh, just put up with it; it’s not that bad. It will be fine’? Or are you out there trying it and saying, ‘Holy cow, this is really inefficient. We have to change this’?”

On the flip side, how much time should be devoted to administrative tasks? The answer depends on the size of your program and the amount of work you have to do, Dr. Faro says. Group directors and department heads normally make themselves available during regular weekday hours. That usually means you’ll have to fit in your clinical work around meetings, budgets, and presentations.

Can You Give Up Clinical Duties?

It’s natural for physicians to be reluctant to relinquish patient care; some reach a point where they have to make the tough decision to stop clinical work altogether.

“You may figure out that you want to pursue an administrative role, but you don’t want to give up clinical work,” says Dr. Gorman, who spent 15 years juggling a full clinical schedule with administrative duties before she became a full-time administrator. “You get plenty of opportunities to make that decision as you’re crossing back and forth.”

You might want to evaluate your options and make the choice sooner rather than later. Once you’re in administration, the decision might be forced upon you. “Eventually, you’ll find that critical things are happening all hours of the day, any given day of the week, in administration as well as clinical practice,” Dr. Faro says. “There’s a point at which you realize that part-time [administrative work] just doesn’t work. You realize that your expertise and guidance are needed.”

Dr. Gorman warns that there are risks and changes involved with becoming a full-time administrator. Once you decide to give up your clinical practice and go the leadership route, your career is “in the hands of someone else,” she points out. “Your position could be eliminated. You could be fired or replaced. … That is a concern. A lot of people keep their hand in on clinical skills for that reason.” You also might find that advancing a management career requires moving to a new organization or a different part of the country.

On the other hand, the rewards of a career in administration can’t be overlooked. “It’s very satisfying personally,” Dr. Faro says. “It’s inventive; you’re constantly solving problems that didn’t exist yesterday.

 

“It’s a different kind of job satisfaction. It’s a very personal decision. There are people who realize that this just isn’t for them.” TH

Jane Jerrard is a freelance writer based in Chicago.

As hospitalists take on more demanding leadership roles, the climb up the career ladder evolves into a juggling act: Hospitalists typically try to handle a full patient load as well as new administrative duties.

If a hospitalist continues to ascend, those administrative duties can begin to consume the schedule. The individual—and the group—could face important decisions about priorities, schedules, and money.

“Hospital medicine is only ten years old; we’re still trying to figure this out,” says Mary Jo Gorman, MD, MBA, chief executive officer of Advanced ICU Care in St. Louis and a past president of SHM. “It’s always a challenge. You identify that you have a need for someone to take charge of an administrative task, but it can take as long as a year to free up [the hospitalist’s] time so that it can get done.”

If you have found yourself in this position, you know that something has to give. “I’ve seen high-energy physicians who think that they can do it all—and they had to,” says Joan C. Faro, MD, FACP, MBA, chief medical officer at John T. Mather Memorial Hospital in Port Jefferson, N.Y. “That is not sustainable. It can’t last forever.”

The question is, how can a hospitalist effectively balance their clinical and administrative duties? Furthermore, what happens when the scales tip in favor—and to the detriment—of one or the other?

When the Juggling Begins

Hospitalists usually add “extra” duties to their normal workloads to advance their careers. Few relinquish their clinical duties as they join committees, further their training, lead a research project, or take on administrative duties.

Dr. Faro says a hospitalist should be able to “head up a focused project or serve on committees” and still be able to meet all their clinical duties. “Once you get beyond that, you need a certain amount of protected time” for administrative or project work, she says. “And when you start to have people reporting to you, you absolutely need that protected time.”

Assigning administrative tasks to physicians who regularly see patients depends on the group structure and requires a clearly defined job description. “If a group is really going to make this work, then you have to pay people for that extra time,” Dr. Gorman says.

Ideally, HM groups have job descriptions for physicians who are called upon to see patients and handle administrative duties. Contracts should include specifications for “protected time,” as well as compensation for new responsibilities.

Clinical-Hour Cutbacks

As administrative duties grow, something has to give. Hospitalists who want to pursue positions of leadership know that that something is hours spent delivering patient care. “If you’re a hospitalist-administrator who wants to make the leap to vice president or department chair or chief medical officer, you need to devote a lot of time to your administrative work,” Dr. Faro says. “You can’t make that leap without putting in those hours.”

So what is a reasonable division of time for, say, the director of an HM program or department? “It’s impossible to pinpoint, but I’d say roughly that [a director] should spend not less than 25% or 30% of their time, and certainly not more than 50% of their time, on clinical work,” Dr. Faro estimates.

 

Even upper-level physician-administrators should maintain a clinical practice simply to monitor the work their department is doing. “It’s not about [clinical] skills as much as it is about whether you can relate to physicians’ day-to-day work, to their frustrations,” Dr. Gorman says. “That’s a management challenge no matter who you are. For example, if hospitalists are complaining about a new EMR [electronic medical record] system, are you going to say, ‘Oh, just put up with it; it’s not that bad. It will be fine’? Or are you out there trying it and saying, ‘Holy cow, this is really inefficient. We have to change this’?”

On the flip side, how much time should be devoted to administrative tasks? The answer depends on the size of your program and the amount of work you have to do, Dr. Faro says. Group directors and department heads normally make themselves available during regular weekday hours. That usually means you’ll have to fit in your clinical work around meetings, budgets, and presentations.

Can You Give Up Clinical Duties?

It’s natural for physicians to be reluctant to relinquish patient care; some reach a point where they have to make the tough decision to stop clinical work altogether.

“You may figure out that you want to pursue an administrative role, but you don’t want to give up clinical work,” says Dr. Gorman, who spent 15 years juggling a full clinical schedule with administrative duties before she became a full-time administrator. “You get plenty of opportunities to make that decision as you’re crossing back and forth.”

You might want to evaluate your options and make the choice sooner rather than later. Once you’re in administration, the decision might be forced upon you. “Eventually, you’ll find that critical things are happening all hours of the day, any given day of the week, in administration as well as clinical practice,” Dr. Faro says. “There’s a point at which you realize that part-time [administrative work] just doesn’t work. You realize that your expertise and guidance are needed.”

Dr. Gorman warns that there are risks and changes involved with becoming a full-time administrator. Once you decide to give up your clinical practice and go the leadership route, your career is “in the hands of someone else,” she points out. “Your position could be eliminated. You could be fired or replaced. … That is a concern. A lot of people keep their hand in on clinical skills for that reason.” You also might find that advancing a management career requires moving to a new organization or a different part of the country.

On the other hand, the rewards of a career in administration can’t be overlooked. “It’s very satisfying personally,” Dr. Faro says. “It’s inventive; you’re constantly solving problems that didn’t exist yesterday.

 

“It’s a different kind of job satisfaction. It’s a very personal decision. There are people who realize that this just isn’t for them.” TH

Jane Jerrard is a freelance writer based in Chicago.

Given the varying access to acute-stroke expertise and the roles hospitalists play in treatment (see “Spotlight on Stroke,” p. 1), stroke protocol differs from hospital to hospital throughout the U.S. One response is known as “drip and ship.” Physicians at remote hospitals consult experts at a tertiary-care medical center by phone or video before initiating clot-busting intravenous recombinant tissue plasminogen activator (t-PA) within its three- to 4.5-hour therapeutic window. Once t-PA is administered, the patient is transferred to the medical center for ongoing care.

“But what is the best way to provide that expertise at the bedside to support the first-responding physician who is not a stroke expert?” asks Lee Schwamm, MD, director of acute-stroke services at Massachusetts General Hospital (MGH) in Boston. While the goal is to disseminate stroke treatment expertise as widely as possible, there are other benefits to the arrangement, from the quality of the infrastructure, ongoing education, and a growing relationship that is more than just “transactional” telemedicine.

MGH and Brigham and Women’s Hospital are the hubs for the relationship-building Partners TeleStroke Network. It connects 27 participating hospitals across three states with an escalating chain of access to stroke resources. Spoke hospitals transmit, through a secure link, such clinical data as noncontrast head CT scans to the hub, where a stroke expert “examines” the patient via live video feed and shares in the responsibility for deciding whether to initiate t-PA. The network’s resources include clinical and information technology advocates at the hub and spokes; managers of business processes, contracts, licensure, and credentialing; consultation recording for quality purposes; regular telemedicine grand rounds; and the network’s leadership in an alliance of hub-and-spokes stroke networks at other academic medical centers. “This is not a game to play casually. It’s about developing new healthcare delivery models, with lots of complicating factors,” Dr. Schwamm says.

Hospitalists should not only note that stroke care is coming under greater regulatory scrutiny, but also that stroke information increasingly is available on the Web, Dr. Schwamm says. He also urges hospitals to participate in one of the national quality programs for stroke care, including the American Stroke Association’s Get with the Guidelines: Stroke, the Joint Commission’s primary stroke center accreditation, or the CDC’s Paul Coverdell National Acute Stroke Registry. “Each of these provides a structure for improving the quality of stroke care,” Dr. Schwamm explains, “and is money well spent by the hospital.”

Given the varying access to acute-stroke expertise and the roles hospitalists play in treatment (see “Spotlight on Stroke,” p. 1), stroke protocol differs from hospital to hospital throughout the U.S. One response is known as “drip and ship.” Physicians at remote hospitals consult experts at a tertiary-care medical center by phone or video before initiating clot-busting intravenous recombinant tissue plasminogen activator (t-PA) within its three- to 4.5-hour therapeutic window. Once t-PA is administered, the patient is transferred to the medical center for ongoing care.

“But what is the best way to provide that expertise at the bedside to support the first-responding physician who is not a stroke expert?” asks Lee Schwamm, MD, director of acute-stroke services at Massachusetts General Hospital (MGH) in Boston. While the goal is to disseminate stroke treatment expertise as widely as possible, there are other benefits to the arrangement, from the quality of the infrastructure, ongoing education, and a growing relationship that is more than just “transactional” telemedicine.

MGH and Brigham and Women’s Hospital are the hubs for the relationship-building Partners TeleStroke Network. It connects 27 participating hospitals across three states with an escalating chain of access to stroke resources. Spoke hospitals transmit, through a secure link, such clinical data as noncontrast head CT scans to the hub, where a stroke expert “examines” the patient via live video feed and shares in the responsibility for deciding whether to initiate t-PA. The network’s resources include clinical and information technology advocates at the hub and spokes; managers of business processes, contracts, licensure, and credentialing; consultation recording for quality purposes; regular telemedicine grand rounds; and the network’s leadership in an alliance of hub-and-spokes stroke networks at other academic medical centers. “This is not a game to play casually. It’s about developing new healthcare delivery models, with lots of complicating factors,” Dr. Schwamm says.

Hospitalists should not only note that stroke care is coming under greater regulatory scrutiny, but also that stroke information increasingly is available on the Web, Dr. Schwamm says. He also urges hospitals to participate in one of the national quality programs for stroke care, including the American Stroke Association’s Get with the Guidelines: Stroke, the Joint Commission’s primary stroke center accreditation, or the CDC’s Paul Coverdell National Acute Stroke Registry. “Each of these provides a structure for improving the quality of stroke care,” Dr. Schwamm explains, “and is money well spent by the hospital.”

Given the varying access to acute-stroke expertise and the roles hospitalists play in treatment (see “Spotlight on Stroke,” p. 1), stroke protocol differs from hospital to hospital throughout the U.S. One response is known as “drip and ship.” Physicians at remote hospitals consult experts at a tertiary-care medical center by phone or video before initiating clot-busting intravenous recombinant tissue plasminogen activator (t-PA) within its three- to 4.5-hour therapeutic window. Once t-PA is administered, the patient is transferred to the medical center for ongoing care.

“But what is the best way to provide that expertise at the bedside to support the first-responding physician who is not a stroke expert?” asks Lee Schwamm, MD, director of acute-stroke services at Massachusetts General Hospital (MGH) in Boston. While the goal is to disseminate stroke treatment expertise as widely as possible, there are other benefits to the arrangement, from the quality of the infrastructure, ongoing education, and a growing relationship that is more than just “transactional” telemedicine.

MGH and Brigham and Women’s Hospital are the hubs for the relationship-building Partners TeleStroke Network. It connects 27 participating hospitals across three states with an escalating chain of access to stroke resources. Spoke hospitals transmit, through a secure link, such clinical data as noncontrast head CT scans to the hub, where a stroke expert “examines” the patient via live video feed and shares in the responsibility for deciding whether to initiate t-PA. The network’s resources include clinical and information technology advocates at the hub and spokes; managers of business processes, contracts, licensure, and credentialing; consultation recording for quality purposes; regular telemedicine grand rounds; and the network’s leadership in an alliance of hub-and-spokes stroke networks at other academic medical centers. “This is not a game to play casually. It’s about developing new healthcare delivery models, with lots of complicating factors,” Dr. Schwamm says.

Hospitalists should not only note that stroke care is coming under greater regulatory scrutiny, but also that stroke information increasingly is available on the Web, Dr. Schwamm says. He also urges hospitals to participate in one of the national quality programs for stroke care, including the American Stroke Association’s Get with the Guidelines: Stroke, the Joint Commission’s primary stroke center accreditation, or the CDC’s Paul Coverdell National Acute Stroke Registry. “Each of these provides a structure for improving the quality of stroke care,” Dr. Schwamm explains, “and is money well spent by the hospital.”

Warfarin (coumadin) differs from most other drugs in that the dosage required to achieve a desired therapeutic effect varies greatly among individuals. This variability can lead to therapeutic failure, potentially resulting in new thrombosis, or, at the other extreme, to life-threatening bleeding.

Further, there is no reliable means to identify patients who require unusually high doses of warfarin, although genetic testing may become available in the future.

See related patient information

Warfarin, a coumarin derivative first synthesized in 1948, is still the only oral anticoagulant available for long-term use in the United States. Indications for its use include the treatment and, to a lesser extent, the prevention of arterial and venous thromboembolism. It is also used for long-term anticoagulation in patients with atrial arrhythmias (atrial fibrillation and atrial flutter) and mechanical heart valves.

In the paragraphs that follow, we review the causes of warfarin resistance and how to recognize and manage it.

WHAT IS WARFARIN RESISTANCE?

Resistance to warfarin has been described as the inability to prolong the prothrombin time or raise the international normalized ratio (INR) into the therapeutic range when the drug is given at normally prescribed doses.¹

However, a higher warfarin requirement does not itself establish the diagnosis of warfarin resistance. The prevalence of warfarin resistance varies by patient population and is difficult to determine. The difficulty lies largely in accounting for dietary factors and in defining normal metabolic variations among individuals.

The range of normally recommended daily or weekly warfarin doses to maintain a therapeutic prothrombin time or INR depends on the study population. Nevertheless, patients who need more than 105 mg per week (15 mg/day) should be considered warfarin-resistant. These patients are likely to be in the top 5% for warfarin doses within an anticoagulated cohort.

Warfarin resistance is different than warfarin failure, which is defined as a new thrombotic event despite a therapeutic prothrombin time and INR. This situation is commonly seen in patients with malignant diseases.

An important characteristic of warfarin resistance is that patients need much smaller doses of vitamin K to reverse the effect of warfarin.² Thijssen³ showed that, in warfarin-resistant rats, warfarin did not irreversibly inhibit vitamin K₁ 2,3-epoxide reductase (VKORC1) activity. This is consistent with the vitamin K hypersensitivity observed in warfarin-resistant people.^2,3

WHAT CAUSES WARFARIN RESISTANCE?

Warfarin resistance can be classified in practical terms as acquired vs hereditary, or in mechanistic terms as pharmacokinetic vs pharmacodynamic.

Acquired vs hereditary resistance

Hulse⁴ categorizes warfarin resistance as either acquired or hereditary.

Acquired resistance to warfarin may result from:

• Poor patient compliance (the most common cause)
• High consumption of vitamin K
• 
  Decreased absorption of warfarin
• Increased clearance (see Warfarin is metabolized by P450 enzymes^5–11)
• Drug interactions (Table 1).^12,13

Hereditary resistance has been postulated to be caused by genetic factors that result either in faster metabolism of the drug (a form of pharmacokinetic resistance) or in lower activity of the drug (pharmacodynamic resistance). Polymorphisms may play a role, as some VKORC1 and CYP2C9 variant alleles are known to be associated with increased sensitivity to warfarin.¹⁴

However, the genetic mechanisms of warfarin resistance are not clearly understood, despite several case reports of hereditary resistance confirmed by similar patterns of resistance in immediate family members.^15–19 More than one mechanism is likely. There is ample room for further insight into genetic polymorphisms underlying hereditary warfarin resistance. More on this topic is included in the sections below.

 

 

Pharmacokinetic resistance

Pharmacokinetic resistance can result from diminished absorption or increased elimination of the drug. Causes of diminished absorption include emesis, diarrhea, and malabsorption syndrome.

The mechanism of increased warfarin clearance has not been delineated, although the following have been implicated.

Genetic factors. Duplication or multiplication of cytochrome P450 enzyme genes has been described as contributing to a phenotype of ultrarapid metabolism. Some people may carry multiple copies of the CYP2C9 gene, as has already been reported for cytochrome P450 CYP2D6 and CYP2A6.^7,8 It is also plausible that rare allelic variants of CYP2C9 exist that are associated with higher-than-normal activity, given that there are alleles known to predispose to warfarin sensitivity.

Hypoalbuminemia may increase the free fraction of warfarin, leading to enhanced rates of clearance and a shorter plasma half-life.¹⁵

Hyperalbuminemia may paradoxically also contribute to warfarin resistance via drug binding.

Hyperlipidemia. Several observers have found that lowering serum lipids, primarily triglycerides, increases the sensitivity to warfarin irrespective of the means used to achieve this decrease.²⁰ This most likely results in a decreased pool of vitamin K, some of which is bound to triglycerides.²¹ Conversely, patients receiving intravenous lipids with total parenteral nutrition have also been diagnosed clinically with warfarin resistance,²² and rat models have shown an association between a lipidrich diet and increased vitamin K-dependent factor activity.²³

Diuretics may decrease the response to warfarin by reducing the plasma volume, with a subsequent increase in clotting factor activity.²⁴

Pharmacodynamic resistance

Potential mechanisms of pharmacodynamic warfarin resistance described in rats and in people include:

• Increased affinity of vitamin K₁, 2,3-epoxide reductase complex (VKOR) for vitamin K^25,26 (see How warfarin works^{2,10,11,27–30})
• Prolongation of normal clotting factor activity¹⁶
• Production of clotting factors that is not dependent on vitamin K¹⁶
• Decreased VKOR sensitivity to warfarin.²⁶

In rats, these mechanisms are manifested by relatively high doses of warfarin being required to achieve poisoning. In humans, they result in high doses being needed to achieve a therapeutic effect in the setting of normal warfarin pharmacokinetics, normal warfarin concentration, and normal half-lives of blood clotting proteins.

Genetics of pharmacodynamic resistance. Pharmacodynamic warfarin resistance has also been described with inheritance of a monogenetic dominant trait. An early study by O’Reilly²⁴ traced anticoagulation resistance to a genetically linked abnormality of interaction between warfarin and a putative vitamin K receptor.

In one patient with hereditary resistance and high warfarin requirements, a heterozygous point mutation in the VKORC1 gene was identified.³¹ This results in a substitution that lies in a conserved (normally constant or unchanging DNA sequence in a genome) region of VKORC1 that contains three of four previously identified amino acid substitutions associated with warfarin resistance (Val29Leu, Val45Ala, and Arg58Gly). Further investigation is required to fully characterize the structure-function relationship for VKORC1 and to determine the relationship between the VKORC1 genotype and other pharmacogenetic determinants of warfarin dose-response.

Separately, Loebstein et al³² reported a new mutation, Asp36Tyr, which was common in Jewish ethnic groups of Ethiopian descent (in whom the prevalence is 5%) and Ashkenazi descent (prevalence 4%). In that study, Asp36Tyr carriers needed doses of more than 70 mg per week, placing them towards the high end of the usual warfarin dosing range.

Daly and Aithal⁷ discovered that warfarinresistant rats overexpressed a protein known as calumenin. This protein is situated in the endoplasmic reticulum and appears to interact with VKOR, decreasing the binding of warfarin. In mice, the calumenin gene is located on chromosome 7, where the gene for VKORC1 is also located.

 

DIAGNOSIS BY HISTORY AND LABORATORY STUDIES

A full drug and diet history is invaluable in diagnosing potential causes of warfarin resistance (Table 1).

Plasma warfarin levels that are subtherapeutic should raise suspicion of intestinal malabsorption or poor compliance. Poor compliance might be more appropriately seen as a mimic of warfarin resistance. Studies in humans suggest that a therapeutic total plasma warfarin level lies between 0.5 μg/mL and 3.0 μg/mL,¹⁰ though the range may vary among laboratories and patient populations.

Warfarin absorption and clearance can be evaluated by analyzing plasma levels at specific intervals after administration, eg, every 60 to 180 minutes. The drug’s half-life can be determined on the basis of its concentrations in different time samples. Normally, the S-enantiomer of warfarin is cleared at twice the rate of the R-enantiomer (5.2 vs 2.5 mL/min/70 kg).⁸ A normal clearance rate confirms that resistance to warfarin is not due to enhanced elimination.

Clotting assays of factors II, VII, IX, and X may be a more precise way to assess the pharmacodynamics of warfarin,¹⁰ although there is no strong evidence to support routine use of such assays. Some studies suggest targeting factor II and factor X activity levels of 10% to 30% of normal biologic activity for a therapeutic warfarin effect in patients with an unreliable or prolonged baseline prothrombin time and INR, such as those with lupus anticoagulant.

An algorithm. Bentley et al³³ suggest using the plasma warfarin level in an algorithm to determine the type of resistance pattern. Plasma warfarin levels are typically measured by regional specialized reference laboratories with a turnaround time of 2 to 7 days, as opposed to 24 hours for factor II and X activity. Our suggested algorithm for evaluation of suspected warfarin resistance is shown in Figure 2.

TREAT THE CAUSE

Once the type of warfarin resistance has been determined, treatment should be oriented toward the cause.

Educate the patient

The importance of compliance should be reinforced. Educating the patient about diet and other medications that may interact with warfarin is also important. (See an example of patient education material.)

Increase the warfarin dose

If the patient truly has hereditary resistance, there are two approaches to treatment.

The first is to increase the warfarin dose until the prothrombin time and INR are in the therapeutic ranges. When indicated, the warfarin dose can be safely titrated upward to more than 100 mg per day in patients who are monitored regularly—as all patients on chronic warfarin therapy should be—and whose other medications are otherwise stable. One such example is reported in a warfarinresistant patient who needed 145 mg/day to maintain a therapeutic prothrombin time.²²

Try other anticoagulants?

The second approach is to change to another type of anticoagulant. However, there is no strong evidence in favor of this approach over prescribing larger dosages of warfarin.

Other anticoagulant drugs currently available in the United States include subcutaneous heparins (unfractionated and low-molecular-weight heparins) and the subcutaneous factor Xa inhibitor fondaparinux (Arixtra).

Agents not available in the United States include the following.

Dabigatran, an oral direct thrombin inhibitor, is undergoing phase 3 studies of its use for long-term anticoagulation.

Rivaroxaban (a direct factor Xa inhibitor) and dabigatran have been approved in Canada and the European Union to prevent venous thromboembolism after knee and hip arthroplasty, based on prospective comparisons with enoxaparin (Lovenox).^34–37

Vitamin K antagonists other than warfarin that are not available in the United States include bishydroxycoumarin (which has limitations including slow absorption and high frequency of gastrointestinal side effects), phenprocoumon, and acenocoumarol. Another is phenindione, which has been associated with serious hypersensitivity reactions, some of which proved fatal and occurred within a few weeks of initiating therapy.

Warfarin (coumadin) differs from most other drugs in that the dosage required to achieve a desired therapeutic effect varies greatly among individuals. This variability can lead to therapeutic failure, potentially resulting in new thrombosis, or, at the other extreme, to life-threatening bleeding.

Further, there is no reliable means to identify patients who require unusually high doses of warfarin, although genetic testing may become available in the future.

See related patient information

Warfarin, a coumarin derivative first synthesized in 1948, is still the only oral anticoagulant available for long-term use in the United States. Indications for its use include the treatment and, to a lesser extent, the prevention of arterial and venous thromboembolism. It is also used for long-term anticoagulation in patients with atrial arrhythmias (atrial fibrillation and atrial flutter) and mechanical heart valves.

In the paragraphs that follow, we review the causes of warfarin resistance and how to recognize and manage it.

WHAT IS WARFARIN RESISTANCE?

Resistance to warfarin has been described as the inability to prolong the prothrombin time or raise the international normalized ratio (INR) into the therapeutic range when the drug is given at normally prescribed doses.¹

However, a higher warfarin requirement does not itself establish the diagnosis of warfarin resistance. The prevalence of warfarin resistance varies by patient population and is difficult to determine. The difficulty lies largely in accounting for dietary factors and in defining normal metabolic variations among individuals.

The range of normally recommended daily or weekly warfarin doses to maintain a therapeutic prothrombin time or INR depends on the study population. Nevertheless, patients who need more than 105 mg per week (15 mg/day) should be considered warfarin-resistant. These patients are likely to be in the top 5% for warfarin doses within an anticoagulated cohort.

Warfarin resistance is different than warfarin failure, which is defined as a new thrombotic event despite a therapeutic prothrombin time and INR. This situation is commonly seen in patients with malignant diseases.

An important characteristic of warfarin resistance is that patients need much smaller doses of vitamin K to reverse the effect of warfarin.² Thijssen³ showed that, in warfarin-resistant rats, warfarin did not irreversibly inhibit vitamin K₁ 2,3-epoxide reductase (VKORC1) activity. This is consistent with the vitamin K hypersensitivity observed in warfarin-resistant people.^2,3

WHAT CAUSES WARFARIN RESISTANCE?

Warfarin resistance can be classified in practical terms as acquired vs hereditary, or in mechanistic terms as pharmacokinetic vs pharmacodynamic.

Acquired vs hereditary resistance

Hulse⁴ categorizes warfarin resistance as either acquired or hereditary.

Acquired resistance to warfarin may result from:

• Poor patient compliance (the most common cause)
• High consumption of vitamin K
• 
  Decreased absorption of warfarin
• Increased clearance (see Warfarin is metabolized by P450 enzymes^5–11)
• Drug interactions (Table 1).^12,13

Hereditary resistance has been postulated to be caused by genetic factors that result either in faster metabolism of the drug (a form of pharmacokinetic resistance) or in lower activity of the drug (pharmacodynamic resistance). Polymorphisms may play a role, as some VKORC1 and CYP2C9 variant alleles are known to be associated with increased sensitivity to warfarin.¹⁴

However, the genetic mechanisms of warfarin resistance are not clearly understood, despite several case reports of hereditary resistance confirmed by similar patterns of resistance in immediate family members.^15–19 More than one mechanism is likely. There is ample room for further insight into genetic polymorphisms underlying hereditary warfarin resistance. More on this topic is included in the sections below.

 

 

Pharmacokinetic resistance

Pharmacokinetic resistance can result from diminished absorption or increased elimination of the drug. Causes of diminished absorption include emesis, diarrhea, and malabsorption syndrome.

The mechanism of increased warfarin clearance has not been delineated, although the following have been implicated.

Genetic factors. Duplication or multiplication of cytochrome P450 enzyme genes has been described as contributing to a phenotype of ultrarapid metabolism. Some people may carry multiple copies of the CYP2C9 gene, as has already been reported for cytochrome P450 CYP2D6 and CYP2A6.^7,8 It is also plausible that rare allelic variants of CYP2C9 exist that are associated with higher-than-normal activity, given that there are alleles known to predispose to warfarin sensitivity.

Hypoalbuminemia may increase the free fraction of warfarin, leading to enhanced rates of clearance and a shorter plasma half-life.¹⁵

Hyperalbuminemia may paradoxically also contribute to warfarin resistance via drug binding.

Hyperlipidemia. Several observers have found that lowering serum lipids, primarily triglycerides, increases the sensitivity to warfarin irrespective of the means used to achieve this decrease.²⁰ This most likely results in a decreased pool of vitamin K, some of which is bound to triglycerides.²¹ Conversely, patients receiving intravenous lipids with total parenteral nutrition have also been diagnosed clinically with warfarin resistance,²² and rat models have shown an association between a lipidrich diet and increased vitamin K-dependent factor activity.²³

Diuretics may decrease the response to warfarin by reducing the plasma volume, with a subsequent increase in clotting factor activity.²⁴

Pharmacodynamic resistance

Potential mechanisms of pharmacodynamic warfarin resistance described in rats and in people include:

• Increased affinity of vitamin K₁, 2,3-epoxide reductase complex (VKOR) for vitamin K^25,26 (see How warfarin works^{2,10,11,27–30})
• Prolongation of normal clotting factor activity¹⁶
• Production of clotting factors that is not dependent on vitamin K¹⁶
• Decreased VKOR sensitivity to warfarin.²⁶

In rats, these mechanisms are manifested by relatively high doses of warfarin being required to achieve poisoning. In humans, they result in high doses being needed to achieve a therapeutic effect in the setting of normal warfarin pharmacokinetics, normal warfarin concentration, and normal half-lives of blood clotting proteins.

Genetics of pharmacodynamic resistance. Pharmacodynamic warfarin resistance has also been described with inheritance of a monogenetic dominant trait. An early study by O’Reilly²⁴ traced anticoagulation resistance to a genetically linked abnormality of interaction between warfarin and a putative vitamin K receptor.

In one patient with hereditary resistance and high warfarin requirements, a heterozygous point mutation in the VKORC1 gene was identified.³¹ This results in a substitution that lies in a conserved (normally constant or unchanging DNA sequence in a genome) region of VKORC1 that contains three of four previously identified amino acid substitutions associated with warfarin resistance (Val29Leu, Val45Ala, and Arg58Gly). Further investigation is required to fully characterize the structure-function relationship for VKORC1 and to determine the relationship between the VKORC1 genotype and other pharmacogenetic determinants of warfarin dose-response.

Separately, Loebstein et al³² reported a new mutation, Asp36Tyr, which was common in Jewish ethnic groups of Ethiopian descent (in whom the prevalence is 5%) and Ashkenazi descent (prevalence 4%). In that study, Asp36Tyr carriers needed doses of more than 70 mg per week, placing them towards the high end of the usual warfarin dosing range.

Daly and Aithal⁷ discovered that warfarinresistant rats overexpressed a protein known as calumenin. This protein is situated in the endoplasmic reticulum and appears to interact with VKOR, decreasing the binding of warfarin. In mice, the calumenin gene is located on chromosome 7, where the gene for VKORC1 is also located.

 

DIAGNOSIS BY HISTORY AND LABORATORY STUDIES

A full drug and diet history is invaluable in diagnosing potential causes of warfarin resistance (Table 1).

Plasma warfarin levels that are subtherapeutic should raise suspicion of intestinal malabsorption or poor compliance. Poor compliance might be more appropriately seen as a mimic of warfarin resistance. Studies in humans suggest that a therapeutic total plasma warfarin level lies between 0.5 μg/mL and 3.0 μg/mL,¹⁰ though the range may vary among laboratories and patient populations.

Warfarin absorption and clearance can be evaluated by analyzing plasma levels at specific intervals after administration, eg, every 60 to 180 minutes. The drug’s half-life can be determined on the basis of its concentrations in different time samples. Normally, the S-enantiomer of warfarin is cleared at twice the rate of the R-enantiomer (5.2 vs 2.5 mL/min/70 kg).⁸ A normal clearance rate confirms that resistance to warfarin is not due to enhanced elimination.

Clotting assays of factors II, VII, IX, and X may be a more precise way to assess the pharmacodynamics of warfarin,¹⁰ although there is no strong evidence to support routine use of such assays. Some studies suggest targeting factor II and factor X activity levels of 10% to 30% of normal biologic activity for a therapeutic warfarin effect in patients with an unreliable or prolonged baseline prothrombin time and INR, such as those with lupus anticoagulant.

An algorithm. Bentley et al³³ suggest using the plasma warfarin level in an algorithm to determine the type of resistance pattern. Plasma warfarin levels are typically measured by regional specialized reference laboratories with a turnaround time of 2 to 7 days, as opposed to 24 hours for factor II and X activity. Our suggested algorithm for evaluation of suspected warfarin resistance is shown in Figure 2.

TREAT THE CAUSE

Once the type of warfarin resistance has been determined, treatment should be oriented toward the cause.

Educate the patient

The importance of compliance should be reinforced. Educating the patient about diet and other medications that may interact with warfarin is also important. (See an example of patient education material.)

Increase the warfarin dose

If the patient truly has hereditary resistance, there are two approaches to treatment.

The first is to increase the warfarin dose until the prothrombin time and INR are in the therapeutic ranges. When indicated, the warfarin dose can be safely titrated upward to more than 100 mg per day in patients who are monitored regularly—as all patients on chronic warfarin therapy should be—and whose other medications are otherwise stable. One such example is reported in a warfarinresistant patient who needed 145 mg/day to maintain a therapeutic prothrombin time.²²

Try other anticoagulants?

The second approach is to change to another type of anticoagulant. However, there is no strong evidence in favor of this approach over prescribing larger dosages of warfarin.

Other anticoagulant drugs currently available in the United States include subcutaneous heparins (unfractionated and low-molecular-weight heparins) and the subcutaneous factor Xa inhibitor fondaparinux (Arixtra).

Agents not available in the United States include the following.

Dabigatran, an oral direct thrombin inhibitor, is undergoing phase 3 studies of its use for long-term anticoagulation.

Rivaroxaban (a direct factor Xa inhibitor) and dabigatran have been approved in Canada and the European Union to prevent venous thromboembolism after knee and hip arthroplasty, based on prospective comparisons with enoxaparin (Lovenox).^34–37

Vitamin K antagonists other than warfarin that are not available in the United States include bishydroxycoumarin (which has limitations including slow absorption and high frequency of gastrointestinal side effects), phenprocoumon, and acenocoumarol. Another is phenindione, which has been associated with serious hypersensitivity reactions, some of which proved fatal and occurred within a few weeks of initiating therapy.

Warfarin (coumadin) differs from most other drugs in that the dosage required to achieve a desired therapeutic effect varies greatly among individuals. This variability can lead to therapeutic failure, potentially resulting in new thrombosis, or, at the other extreme, to life-threatening bleeding.

Further, there is no reliable means to identify patients who require unusually high doses of warfarin, although genetic testing may become available in the future.

See related patient information

Warfarin, a coumarin derivative first synthesized in 1948, is still the only oral anticoagulant available for long-term use in the United States. Indications for its use include the treatment and, to a lesser extent, the prevention of arterial and venous thromboembolism. It is also used for long-term anticoagulation in patients with atrial arrhythmias (atrial fibrillation and atrial flutter) and mechanical heart valves.

In the paragraphs that follow, we review the causes of warfarin resistance and how to recognize and manage it.

WHAT IS WARFARIN RESISTANCE?

Resistance to warfarin has been described as the inability to prolong the prothrombin time or raise the international normalized ratio (INR) into the therapeutic range when the drug is given at normally prescribed doses.¹

However, a higher warfarin requirement does not itself establish the diagnosis of warfarin resistance. The prevalence of warfarin resistance varies by patient population and is difficult to determine. The difficulty lies largely in accounting for dietary factors and in defining normal metabolic variations among individuals.

The range of normally recommended daily or weekly warfarin doses to maintain a therapeutic prothrombin time or INR depends on the study population. Nevertheless, patients who need more than 105 mg per week (15 mg/day) should be considered warfarin-resistant. These patients are likely to be in the top 5% for warfarin doses within an anticoagulated cohort.

Warfarin resistance is different than warfarin failure, which is defined as a new thrombotic event despite a therapeutic prothrombin time and INR. This situation is commonly seen in patients with malignant diseases.

An important characteristic of warfarin resistance is that patients need much smaller doses of vitamin K to reverse the effect of warfarin.² Thijssen³ showed that, in warfarin-resistant rats, warfarin did not irreversibly inhibit vitamin K₁ 2,3-epoxide reductase (VKORC1) activity. This is consistent with the vitamin K hypersensitivity observed in warfarin-resistant people.^2,3

WHAT CAUSES WARFARIN RESISTANCE?

Warfarin resistance can be classified in practical terms as acquired vs hereditary, or in mechanistic terms as pharmacokinetic vs pharmacodynamic.

Acquired vs hereditary resistance

Hulse⁴ categorizes warfarin resistance as either acquired or hereditary.

Acquired resistance to warfarin may result from:

• Poor patient compliance (the most common cause)
• High consumption of vitamin K
• 
  Decreased absorption of warfarin
• Increased clearance (see Warfarin is metabolized by P450 enzymes^5–11)
• Drug interactions (Table 1).^12,13

Hereditary resistance has been postulated to be caused by genetic factors that result either in faster metabolism of the drug (a form of pharmacokinetic resistance) or in lower activity of the drug (pharmacodynamic resistance). Polymorphisms may play a role, as some VKORC1 and CYP2C9 variant alleles are known to be associated with increased sensitivity to warfarin.¹⁴

However, the genetic mechanisms of warfarin resistance are not clearly understood, despite several case reports of hereditary resistance confirmed by similar patterns of resistance in immediate family members.^15–19 More than one mechanism is likely. There is ample room for further insight into genetic polymorphisms underlying hereditary warfarin resistance. More on this topic is included in the sections below.

 

 

Pharmacokinetic resistance

Pharmacokinetic resistance can result from diminished absorption or increased elimination of the drug. Causes of diminished absorption include emesis, diarrhea, and malabsorption syndrome.

The mechanism of increased warfarin clearance has not been delineated, although the following have been implicated.

Genetic factors. Duplication or multiplication of cytochrome P450 enzyme genes has been described as contributing to a phenotype of ultrarapid metabolism. Some people may carry multiple copies of the CYP2C9 gene, as has already been reported for cytochrome P450 CYP2D6 and CYP2A6.^7,8 It is also plausible that rare allelic variants of CYP2C9 exist that are associated with higher-than-normal activity, given that there are alleles known to predispose to warfarin sensitivity.

Hypoalbuminemia may increase the free fraction of warfarin, leading to enhanced rates of clearance and a shorter plasma half-life.¹⁵

Hyperalbuminemia may paradoxically also contribute to warfarin resistance via drug binding.

Hyperlipidemia. Several observers have found that lowering serum lipids, primarily triglycerides, increases the sensitivity to warfarin irrespective of the means used to achieve this decrease.²⁰ This most likely results in a decreased pool of vitamin K, some of which is bound to triglycerides.²¹ Conversely, patients receiving intravenous lipids with total parenteral nutrition have also been diagnosed clinically with warfarin resistance,²² and rat models have shown an association between a lipidrich diet and increased vitamin K-dependent factor activity.²³

Diuretics may decrease the response to warfarin by reducing the plasma volume, with a subsequent increase in clotting factor activity.²⁴

Pharmacodynamic resistance

Potential mechanisms of pharmacodynamic warfarin resistance described in rats and in people include:

• Increased affinity of vitamin K₁, 2,3-epoxide reductase complex (VKOR) for vitamin K^25,26 (see How warfarin works^{2,10,11,27–30})
• Prolongation of normal clotting factor activity¹⁶
• Production of clotting factors that is not dependent on vitamin K¹⁶
• Decreased VKOR sensitivity to warfarin.²⁶

In rats, these mechanisms are manifested by relatively high doses of warfarin being required to achieve poisoning. In humans, they result in high doses being needed to achieve a therapeutic effect in the setting of normal warfarin pharmacokinetics, normal warfarin concentration, and normal half-lives of blood clotting proteins.

Genetics of pharmacodynamic resistance. Pharmacodynamic warfarin resistance has also been described with inheritance of a monogenetic dominant trait. An early study by O’Reilly²⁴ traced anticoagulation resistance to a genetically linked abnormality of interaction between warfarin and a putative vitamin K receptor.

In one patient with hereditary resistance and high warfarin requirements, a heterozygous point mutation in the VKORC1 gene was identified.³¹ This results in a substitution that lies in a conserved (normally constant or unchanging DNA sequence in a genome) region of VKORC1 that contains three of four previously identified amino acid substitutions associated with warfarin resistance (Val29Leu, Val45Ala, and Arg58Gly). Further investigation is required to fully characterize the structure-function relationship for VKORC1 and to determine the relationship between the VKORC1 genotype and other pharmacogenetic determinants of warfarin dose-response.

Separately, Loebstein et al³² reported a new mutation, Asp36Tyr, which was common in Jewish ethnic groups of Ethiopian descent (in whom the prevalence is 5%) and Ashkenazi descent (prevalence 4%). In that study, Asp36Tyr carriers needed doses of more than 70 mg per week, placing them towards the high end of the usual warfarin dosing range.

Daly and Aithal⁷ discovered that warfarinresistant rats overexpressed a protein known as calumenin. This protein is situated in the endoplasmic reticulum and appears to interact with VKOR, decreasing the binding of warfarin. In mice, the calumenin gene is located on chromosome 7, where the gene for VKORC1 is also located.

 

DIAGNOSIS BY HISTORY AND LABORATORY STUDIES

A full drug and diet history is invaluable in diagnosing potential causes of warfarin resistance (Table 1).

Plasma warfarin levels that are subtherapeutic should raise suspicion of intestinal malabsorption or poor compliance. Poor compliance might be more appropriately seen as a mimic of warfarin resistance. Studies in humans suggest that a therapeutic total plasma warfarin level lies between 0.5 μg/mL and 3.0 μg/mL,¹⁰ though the range may vary among laboratories and patient populations.

Warfarin absorption and clearance can be evaluated by analyzing plasma levels at specific intervals after administration, eg, every 60 to 180 minutes. The drug’s half-life can be determined on the basis of its concentrations in different time samples. Normally, the S-enantiomer of warfarin is cleared at twice the rate of the R-enantiomer (5.2 vs 2.5 mL/min/70 kg).⁸ A normal clearance rate confirms that resistance to warfarin is not due to enhanced elimination.

Clotting assays of factors II, VII, IX, and X may be a more precise way to assess the pharmacodynamics of warfarin,¹⁰ although there is no strong evidence to support routine use of such assays. Some studies suggest targeting factor II and factor X activity levels of 10% to 30% of normal biologic activity for a therapeutic warfarin effect in patients with an unreliable or prolonged baseline prothrombin time and INR, such as those with lupus anticoagulant.

An algorithm. Bentley et al³³ suggest using the plasma warfarin level in an algorithm to determine the type of resistance pattern. Plasma warfarin levels are typically measured by regional specialized reference laboratories with a turnaround time of 2 to 7 days, as opposed to 24 hours for factor II and X activity. Our suggested algorithm for evaluation of suspected warfarin resistance is shown in Figure 2.

TREAT THE CAUSE

Once the type of warfarin resistance has been determined, treatment should be oriented toward the cause.

Educate the patient

The importance of compliance should be reinforced. Educating the patient about diet and other medications that may interact with warfarin is also important. (See an example of patient education material.)

Increase the warfarin dose

If the patient truly has hereditary resistance, there are two approaches to treatment.

The first is to increase the warfarin dose until the prothrombin time and INR are in the therapeutic ranges. When indicated, the warfarin dose can be safely titrated upward to more than 100 mg per day in patients who are monitored regularly—as all patients on chronic warfarin therapy should be—and whose other medications are otherwise stable. One such example is reported in a warfarinresistant patient who needed 145 mg/day to maintain a therapeutic prothrombin time.²²

Try other anticoagulants?

The second approach is to change to another type of anticoagulant. However, there is no strong evidence in favor of this approach over prescribing larger dosages of warfarin.

Other anticoagulant drugs currently available in the United States include subcutaneous heparins (unfractionated and low-molecular-weight heparins) and the subcutaneous factor Xa inhibitor fondaparinux (Arixtra).

Agents not available in the United States include the following.

Dabigatran, an oral direct thrombin inhibitor, is undergoing phase 3 studies of its use for long-term anticoagulation.

Rivaroxaban (a direct factor Xa inhibitor) and dabigatran have been approved in Canada and the European Union to prevent venous thromboembolism after knee and hip arthroplasty, based on prospective comparisons with enoxaparin (Lovenox).^34–37

Vitamin K antagonists other than warfarin that are not available in the United States include bishydroxycoumarin (which has limitations including slow absorption and high frequency of gastrointestinal side effects), phenprocoumon, and acenocoumarol. Another is phenindione, which has been associated with serious hypersensitivity reactions, some of which proved fatal and occurred within a few weeks of initiating therapy.

Prasugrel (Effient) is more potent and consistent in its effects than clopidogrel (Plavix), thus preventing more thrombotic events—but at a price of more bleeding. Therefore, the drugs must be appropriately selected for the individual patient.

Over the last 9 years, the thienopyridines—ticlopidine (Ticlid), clopidogrel, and now prasugrel—have become essential tools for treating acute coronary syndromes.

The usual underlying mechanism of acute coronary syndromes is thrombosis, caused by rupture of atherosclerotic plaque.¹ Accordingly, antithrombotic agents—aspirin, heparin, lowmolecular-weight heparin, glycoprotein IIb/IIIa inhibitors, the direct thrombin inhibitor bivalirudin (Angiomax), and thienopyridines—have all been shown to reduce the risk of major adverse cardiac outcomes in this setting.

In this article, we review the pharmacology and evidence of effectiveness of the thienopyridine drugs, focusing on prasugrel, the latest thienopyridine to be approved by the US Food and Drug Administration (FDA).

THIENOPYRIDINES INHIBIT PLATELET ACTIVATION AND AGGREGATION

Thienopyridines are prodrugs that require conversion by hepatic cytochrome P450 enzymes. The active metabolites bind irreversibly to platelet P2Y12 receptors. Consequently, they permanently block signalling mediated by platelet adenosine diphosphate-P2Y12 receptors, thereby inhibiting glycoprotein IIb/IIIa receptor activation and platelet aggregation.

Aspirin, in contrast, inhibits platelets by blocking the thromboxane-mediated pathway. Therefore, the combination of aspirin plus a thienopyridine has an additive effect.²

The effect of thienopyridines on platelets is irreversible. Therefore, although the half-life of prasugrel’s active metabolite is 3.7 hours, its inhibitory effects last for 96 hours, essentially the time for half the body’s circulating platelets to be replaced.

TICLOPIDINE, THE FIRST THIENOPYRIDINE

Ticlopidine was the first thienopyridine to be approved by the FDA. Its initial studies in unstable angina were small, their designs did not call for patients to concurrently receive aspirin, and all they showed was that ticlopidine was about as beneficial as aspirin. Consequently, the studies had little impact on clinical practice.³

In a pivotal trial,⁴ patients who received coronary stents were randomized to afterward receive either the combination of ticlopidine plus aspirin or anticoagulation therapy with heparin, phenprocoumon (a coumarin derivative available in Europe), and aspirin. At 30 days, an ischemic complication (death, myocardial infarction [MI], repeat intervention) had occurred in 6.2% of the anticoagulation therapy group vs 1.6% of the ticlopidine group, a risk reduction of 75%. Rates of stent occlusion, MI, and revascularization were 80% to 85% lower in the ticlodipine group. This study paved the way for widespread use of thienopyridines.

Ticlopidine’s use was limited, however, by a 2.4% incidence of serious granulocytopenia and rare cases of thrombocytopenic purpura.

BENEFIT OF CLOPIDOGREL

Although prasugrel is the focus of this review, the trials of prasugrel all compared its efficacy with that of clopidogrel. Furthermore, many patients should still receive clopidogrel and not prasugrel, so it is important to be familiar with the evidence of clopidogrel’s benefit.

Once approved for clinical use, clopidogrel was substituted for ticlopidine in patients undergoing coronary stenting on the basis of studies showing it to be at least as effective as ticlopidine and more tolerable. A series of trials of clopidogrel were done in patients across a spectrum of risk groups, from those at high risk of coronary heart disease to those presenting with ST-elevation MI. The time of pretreatment in the studies ranged from 3 hours to 6 days before percutaneous coronary intervention, and the duration of treatment following intervention ranged from 30 days to 1 year.

Clopidogrel in non-ST-elevation acute coronary syndromes

The CURE trial² (Clopidogrel in Unstable Angina to Prevent Recurrent Events), published in 2001, established clopidogrel as a therapy for unstable ischemic syndromes, whether treated medically or with revascularization. In that trial, 12,562 patients with acute coronary syndromes without ST elevation (ie, unstable angina or non-ST-elevation MI), as defined by electrocardiographic changes or positive cardiac markers, were randomized to receive clopidogrel (a 300-mg loading dose followed by 75-mg maintenance doses) or placebo for a mean duration of 9 months. All patients also received aspirin 75 mg to 325 mg daily.

The composite outcome of death from cardiovascular causes, nonfatal MI, or stroke occurred in 20% fewer patients treated with clopidogrel than with placebo (9.3% vs 11.4%). The benefit was similar in patients undergoing revascularization compared with those treated medically.

Although there were significantly more cases of major bleeding in the clopidogrel group than in the placebo group (3.7% vs 2.7%), the number of episodes of life-threatening bleeding or hemorrhagic strokes was the same.

PCI-CURE⁵ was a substudy of the CURE trial in patients who underwent a percutaneous coronary intervention. Patients were pretreated with clopidogrel or placebo for a mean of 6 days before the procedure. Afterward, they all received clopidogrel plus aspirin in an unblinded fashion for 2 to 4 weeks, and then the randomized study drug was resumed for a mean of 8 months.

Significantly fewer adverse events occurred in the clopidogrel group as tallied at the time of the intervention, 1 month later, and 8 months later.

 

 

Clopidogrel in ST-elevation acute MI

The CLARITY-TIMI 28 trial⁶ (Clopidogrel as Adjunctive Reperfusion Therapy—Thrombolysis in Myocardial Infarction 28) showed that adding clopidogrel (a 300-mg loading dose, then 75 mg daily) to aspirin benefitted patients with ST-elevation MI receiving fibrinolytic therapy. At 30 days, cardiovascular death, recurrent MI, or urgent revascularization had occurred in 11.6% of the clopidogrel group vs 14.1% of the placebo group, a statistically significant difference. The rates of major or minor bleeding were no higher in the clopidogrel group than in the placebo group, an especially remarkable finding in patients receiving thrombolytic therapy.

PCI-CLARITY.⁷ About half of the patients in the CLARITY trial ultimately underwent a percutaneous coronary intervention after fibrinolytic therapy, with results reported as the PCI-CLARITY substudy. Like those in PCI-CURE, these patients were randomized to receive pretreatment with either clopidogrel or placebo before the procedure, in this study for a median of 3 days. Both groups received clopidogrel afterward. At 30 days from randomization, the outcome of cardiovascular death, MI, or stroke had occurred in 7.5% of the clopidogrel group compared with 12.0% of the placebo group, which was statistically significant, without any significant excess in the rates of major or minor bleeding.

COMMIT⁸ (the Clopidogrel and Metoprolol in Myocardial Infarction Trial) also showed clopidogrel to be beneficial in patients with acute MI. This trial included more than 45,000 patients in China with acute MI, 93% of whom had ST-segment elevation. In contrast to CLARITY, in COMMIT barely more than half of the patients received fibrinolysis, fewer than 5% proceeded to percutaneous interventions, and no loading dose was given: patients in the clopidogrel group received 75 mg/day from the outset.

At 15 days, the incidence of death, reinfarction, or stroke was 9.2% with clopidogrel compared with 10.1% with placebo, a small but statistically significant difference. Again, the rate of major bleeding was not significantly higher, either overall or in patients over age 70.

Of note, patients over age 75 were excluded from CLARITY, and as mentioned, no loading dose was used in COMMIT. Thus, for patients receiving fibrinolysis who are over age 75, there is no evidence to support the safety of a loading dose, and clopidogrel should be started at 75 mg daily.

Clopidogrel in elective percutaneous coronary intervention

The CREDO trial⁹ (Clopidogrel for the Reduction of Events During Observation) was in patients referred for elective percutaneous coronary intervention. Three to 24 hours before the procedure, the patients received either a 300-mg loading dose of clopidogrel or placebo; afterward, all patients received clopidogrel 75 mg/day for 28 days. All patients also received aspirin.

A clopidogrel loading dose 3 to 24 hours before the intervention did not produce a statistically significant reduction in ischemic events, although a post hoc subgroup analysis suggested that patients who received the loading dose between 6 and 24 hours before did benefit, with a relative risk reduction of 38.6% in the composite end point (P = .051).

After 28 days, the patients who had received the clopidogrel loading dose were continued on clopidogrel, while those in the placebo group were switched back to placebo. At 1 year, the investigators found a significantly lower rate of the composite end point with the prolonged course of clopidogrel (8.5% vs 11.5%).

In summary, these studies found clopidogrel to be beneficial in a broad spectrum of coronary diseases. Subgroup analyses suggest that pretreatment before percutaneous coronary intervention provides additional benefit, particularly if clopidogrel is given at least 6 hours in advance (the time necessary for clopidogrel to cause substantial platelet inhibition).

SOME PATIENTS RESPOND LESS TO CLOPIDOGREL

The level of platelet inhibition induced by clopidogrel varies. In different studies, the frequency of clopidogrel “nonresponsiveness” ranged from 5% to 56% of patients, depending on which test and which cutoff values were used. The distribution of responses to clopidogrel is wide and fits a normal gaussian curve.¹⁰

A large fraction of the population carries a gene that may account for some of the interpatient variation in platelet inhibition with clopidogrel. Carriers of a reduced-function CYP2C19 allele—approximately 30% of people in one study—have significantly lower levels of the active metabolite of clopidogrel, less platelet inhibition from clopidogrel therapy, and a 53% higher rate of death from cardiovascular causes, MI, or stroke.¹¹

 

PRASUGREL, THE NEWEST THIENOPYRIDINE

Prasugrel, FDA-approved in July 2009 for the treatment of acute coronary syndromes, is given in an oral loading dose of 60 mg followed by an oral maintenance dose of 10 mg daily.

Pharmacology of prasugrel vs clopidogrel

As noted previously, the thienopyridines are prodrugs that require hepatic conversion to exert antiplatelet effects.

Metabolism. Prasugrel’s hepatic activation involves a single step, in contrast to the multiple-step process required for activation of clopidogrel. Clopidogrel is primarily hydrolyzed by intestinal and plasma esterases to an inactive terminal metabolite, with the residual unhydrolized drug undergoing a two-step metabolism that depends on cytochrome P450 enzymes. Prasugrel is also extensively hydrolyzed by these esterases, but the intermediate product is then metabolized in a single step to the active sulfhydryl compound, mainly by CYP3A4 and CYP2B6.

Thus, about 80% of an orally absorbed dose of prasugrel is converted to active drug, compared with only 10% to 20% of absorbed clopidogrel.

Time to peak effect. With clopidogrel, maximal inhibition of platelet aggregation occurs 3 to 5 days after starting therapy with 75 mg daily without a loading dose, but within 4 to 6 hours if a loading dose of 300 to 600 mg is given. In contrast, a prasugrel loading dose produces more than 80% of its platelet inhibitory effects by 30 minutes, and peak activity is observed within 4 hours.¹² The platelet inhibition induced by prasugrel at 30 minutes after administration is comparable to the peak effect of clopidogrel at 6 hours.¹³

Dose-response. Prasugrel’s inhibition of platelet aggregation is dose-related.

Prasugrel is about 10 times more potent than clopidogrel and 100 times more potent than ticlopidine. Thus, treatment with 5 mg of prasugrel results in inhibition of platelet activity (distributed in a gaussian curve) very similar to that produced by 75 mg of clopidogrel. On the other hand, even a maintenance dose of 150 mg of clopidogrel inhibits platelet activity to a lesser degree than 10 mg of prasugrel (46% vs 61%),¹⁴ so clopidogrel appears to reach a plateau of platelet inhibition that prasugrel can overcome.

At the approved dose of prasugrel, inhibition of platelet aggregation is significantly greater and there are fewer “nonresponders” than with clopidogrel.

Interactions. Drugs that inhibit CYP3A4 do not inhibit the efficacy of prasugrel, but they can inhibit that of clopidogrel. Some commonly used drugs that have this effect are the statins (eg, atorvastain [Lipitor]) and the macrolide antibiotics (eg, erythromycin). Furthermore, whereas proton pump inhibitors have been shown to diminish the effect of clopidogrel by reducing the formation of its active metabolite, no such effect has been noted with prasugrel.

Prasugrel in phase 2 trials: Finding the optimal dosage

A phase 2 trial compared three prasugrel regimens (loading dose/daily maintenance dose of 40 mg/7.5 mg, 60 mg/10 mg, and 60 mg/15 mg) and standard clopidogrel therapy (300 mg/75 mg) in patients undergoing elective or urgent percutaneous coronary intervention.¹⁵ No significant difference in outcomes was seen in the groups receiving the three prasugrel regimens. However, more “minimal bleeding events” (defined by the criteria of the TIMI trial¹⁶) occurred with high-dose prasugrel than with lower-dose prasugrel or with clopidogrel, leading to use of the intermediate-dose prasugrel regimen (60-mg loading dose, 10-mg daily maintenance) for later trials.

Another phase 2 trial randomized 201 patients undergoing elective percutaneous coronary intervention to receive prasugrel 60 mg/10 mg or clopidogrel 600 mg/150 mg.¹⁴ In all patients, the loading dose was given about 1 hour before cardiac catheterization. As soon as 30 minutes after the loading dose, platelet inhibition was superior with prasugrel (31% vs 5% inhibition of platelet aggregation), and it remained significantly higher at 6 hours (75% vs 32%) and during the maintenance phase (61% vs 46%).

 

 

Phase 3 trial of prasugrel vs clopidogrel: TRITON-TIMI 38

Only one large phase 3 trial of prasugrel has been completed: TRITON-TIMI 38 (the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel—Thrombolysis in Myocardial Infarction),¹⁷ which enrolled adults with moderate-risk to high-risk acute coronary syndromes scheduled to undergo a percutaneous coronary intervention. In this trial, 10,074 patients were enrolled who had moderate-to high-risk unstable angina or non-ST-elevation MI, and 3,534 patients were enrolled who had ST-elevation MI.

Patients were randomized to receive prasugrel (a 60-mg loading dose, then 10 mg daily) or clopidogrel (a 300-mg loading dose, then 75 mg daily) and were treated for 6 to 15 months. All patients also received aspirin.

The primary end point, a composite of death from cardiovascular causes, nonfatal MI, or nonfatal stroke, occurred in significantly fewer patients treated with prasugrel than with clopidogrel (9.9% vs 12.1%, P < .001) (Table 1). Most of the benefit was due to fewer nonfatal MIs during the follow-up period (7.4% vs 9.7%, P < .001). Additionally, the prasugrel group had a significantly lower rate of stent thrombosis compared with the clopidogrel group (1.1% vs 2.4%; P < .001).

These benefits came at a price of more bleeding. Of those patients who did not undergo coronary artery bypass grafting, more experienced bleeding in the prasugrel group than in the clopidogrel group (2.4% vs 1.8%, P = .03), including a higher rate of life-threatening bleeding (1.4% vs 0.89%, P = .01) and fatal bleeding (0.4% vs 0.1%, P = .002). More patients discontinued prasugrel because of hemorrhage (2.5% vs 1.4%, P < .001). In patients who proceeded to coronary artery bypass grafting, the rate of major bleeding was more than four times higher in those who received prasugrel than in those who received clopidogrel (13.4% vs 3.2%, P < .001).

A higher rate of adverse events related to colon cancer was also noted in patients treated with prasugrel, although the authors suggest this may have resulted from the stronger antiplatelet effects of prasugrel bringing more tumors to medical attention due to bleeding.

Overall death rates did not differ significantly between the treatment groups.

In a post hoc analysis,¹⁸ prasugrel was superior to clopidogrel in preventing ischemic events both during the first 3 days following randomization (the “loading phase”) and for the remainder of the trial (the “maintenance phase”). Whereas bleeding risk was similar with the two drugs during the loading phase, prasugrel was subsequently associated with more bleeding during the maintenance phase.

Certain patient subgroups had no net benefit or even suffered harm from prasugrel compared with clopidogrel.¹⁷ Patients with previous stroke or transient ischemic attack had net harm from prasugrel (hazard ratio 1.54, P = .04) and showed a strong trend toward a greater rate of major bleeding (P = .06). Patients age 75 and older and those weighing less than 60 kg had no net benefit from prasugrel.

Cost of prasugrel

Prasugrel is currently priced at 18% more than clopidogrel, with average wholesale prices per pill of $6.65 for prasugrel 10 mg compared with $5.63 for clopidogrel 75 mg. (Prasugrel 10-mg pills cost $6.33 at drugstore.com or $7.60 at CVS; clopidogrel 75-mg pills cost $5.33 at drugstore.com or $6.43 at CVS.) The patent on clopidogrel expires in November 2011, after which the price differential is expected to become significantly greater.

TICAGRELOR, A REVERSIBLE ORAL AGENT

Ticagrelor, the first reversible oral P2Y12 receptor antagonist, is an alternative to thienopyridine therapy for acute coronary syndromes.

Ticagrelor is quickly absorbed, does not require metabolic activation, and has a rapid antiplatelet effect and offset of effect, which closely follow drug-exposure levels. In a large randomized controlled trial in patients with acute coronary syndromes with or without STsegment elevation, treatment with ticagrelor compared with clopidogrel resulted in a significant reduction in death from vascular causes, MI, or stroke (9.8% vs 11.7%).¹⁹

Given its reversible effect on platelet inhibition, ticagrelor may be preferred in patients whose coronary anatomy is unknown and for whom coronary artery bypass grafting is deemed probable. It is still undergoing trials and is not yet approved.

 

TAKE-HOME POINTS

Prasugrel is more potent, more rapid in onset, and more consistent in inhibiting platelet aggregation than clopidogrel. A large clinical trial¹⁷ found prasugrel to be superior to clopidogrel for patients with moderate-to high-risk acute coronary syndromes with high probability of undergoing a percutaneous coronary intervention.

Who should receive prasugrel, and how?

Prasugrel should be given after angiography to patients with non-ST-elevation acute coronary syndromes or at presentation to patients with ST-elevation MI. When used for planned percutaneous coronary intervention, prasugrel should be given at least 30 minutes before the intervention, as was done in phase 2 trials (although its routine use in this situation is not recommended—see below).

It is given in a one-time loading dose of 60 mg by mouth and then maintained with 10 mg by mouth once daily for at least 1 year. (At least 9 months of treatment with a thienopyridine is indicated for patients with acute coronary syndromes who are medically treated, and at least 1 year is indicated following urgent or elective percutaneous coronary intervention, including balloon angioplasty and placement of a bare-metal or drug-eluting stent.)

Who should not receive prasugrel?

For now, prasugrel should be avoided in favor of clopidogrel in patients at higher risk of bleeding. It is clearly contraindicated in patients with prior transient ischemic attack or stroke, for whom the risk of serious bleeding seems to be prohibitive. It should generally be avoided in patients age 75 and older, although it might be considered in those at particularly high risk of stent thrombosis, such as those with diabetes or prior MI. In patients weighing less than 60 kg, the package insert advises a reduced dose (5 mg), although clinical evidence for this practice is lacking.

As yet, we have no data assuring that prasugrel is safe to use in combination with fibrinolytic agents, so patients on thrombolytic therapy for acute MI should continue to receive clopidogrel starting immediately after lysis. Furthermore, in patients who proceeded to coronary artery bypass grafting, the rate of major bleeding was more than four times higher in the prasugrel group than in the clopidogrel group in the TRITON-TIMI 38 trial.¹⁷ No thienopyridine should be given to patients likely to proceed to coronary artery bypass grafting.

Only clopidogrel has evidence supporting its use as an alternative to aspirin for patients with atherosclerotic disease who cannot tolerate aspirin. Neither drug has evidence for use for primary prevention.

Other areas of uncertainty

Prior to angiography. Indications for prasugrel are currently limited by the narrow scope of the trial data. TRITON-TIMI 38,¹⁷ the only large trial completed to date, randomized patients to receive prasugrel only after their coronary anatomy was known, except for ST-elevation MI patients. It is unknown whether the benefits of prasugrel will outweigh the higher risk of bleeding in patients with acute coronary syndromes who do not proceed to percutaneous coronary interventions.

A clinical trial is currently under way comparing prasugrel with clopidogrel in 10,000 patients with acute coronary syndromes who will be medically managed without planned revascularization: A Comparison of Prasugrel and Clopidogrel in Acute Coronary Syndrome Subjects (TRILOGY ACS), ClinicalTrials.gov Identifier: NCT00699998. The trial has an estimated completion date of March 2011.

In cases of non-ST-elevation acute coronary syndrome, it is reasonable to wait to give a thienopyridine until after the coronary anatomy has been defined, if angiography will be completed soon after presentation. For example, a 1-hour delay before giving prasugrel still delivers antiplatelet therapy more quickly than giving clopidogrel on presentation. If longer delays are expected before angiography, however, the patient should be given a loading dose of clopidogrel “up front,” in accordance with guidelines published by the American College of Cardiology, American Heart Association, and European Society of Cardiology,²⁰ which recommend starting a thienopyridine early during hospitalization based on trial data with clopidogrel.

Patients undergoing elective percutaneous coronary intervention are at lower risk of stent thrombosis and other ischemic complications, so it is possible that the benefits of prasugrel would not outweigh the risks in these patients. Thus, prasugrel cannot yet be recommended for routine elective percutaneous coronary intervention except in individual cases in which the interventionalist feels that the patient may be at higher risk of thrombosis.

Prasugrel (Effient) is more potent and consistent in its effects than clopidogrel (Plavix), thus preventing more thrombotic events—but at a price of more bleeding. Therefore, the drugs must be appropriately selected for the individual patient.

Over the last 9 years, the thienopyridines—ticlopidine (Ticlid), clopidogrel, and now prasugrel—have become essential tools for treating acute coronary syndromes.

The usual underlying mechanism of acute coronary syndromes is thrombosis, caused by rupture of atherosclerotic plaque.¹ Accordingly, antithrombotic agents—aspirin, heparin, lowmolecular-weight heparin, glycoprotein IIb/IIIa inhibitors, the direct thrombin inhibitor bivalirudin (Angiomax), and thienopyridines—have all been shown to reduce the risk of major adverse cardiac outcomes in this setting.

In this article, we review the pharmacology and evidence of effectiveness of the thienopyridine drugs, focusing on prasugrel, the latest thienopyridine to be approved by the US Food and Drug Administration (FDA).

THIENOPYRIDINES INHIBIT PLATELET ACTIVATION AND AGGREGATION

Thienopyridines are prodrugs that require conversion by hepatic cytochrome P450 enzymes. The active metabolites bind irreversibly to platelet P2Y12 receptors. Consequently, they permanently block signalling mediated by platelet adenosine diphosphate-P2Y12 receptors, thereby inhibiting glycoprotein IIb/IIIa receptor activation and platelet aggregation.

Aspirin, in contrast, inhibits platelets by blocking the thromboxane-mediated pathway. Therefore, the combination of aspirin plus a thienopyridine has an additive effect.²

The effect of thienopyridines on platelets is irreversible. Therefore, although the half-life of prasugrel’s active metabolite is 3.7 hours, its inhibitory effects last for 96 hours, essentially the time for half the body’s circulating platelets to be replaced.

TICLOPIDINE, THE FIRST THIENOPYRIDINE

Ticlopidine was the first thienopyridine to be approved by the FDA. Its initial studies in unstable angina were small, their designs did not call for patients to concurrently receive aspirin, and all they showed was that ticlopidine was about as beneficial as aspirin. Consequently, the studies had little impact on clinical practice.³

In a pivotal trial,⁴ patients who received coronary stents were randomized to afterward receive either the combination of ticlopidine plus aspirin or anticoagulation therapy with heparin, phenprocoumon (a coumarin derivative available in Europe), and aspirin. At 30 days, an ischemic complication (death, myocardial infarction [MI], repeat intervention) had occurred in 6.2% of the anticoagulation therapy group vs 1.6% of the ticlopidine group, a risk reduction of 75%. Rates of stent occlusion, MI, and revascularization were 80% to 85% lower in the ticlodipine group. This study paved the way for widespread use of thienopyridines.

Ticlopidine’s use was limited, however, by a 2.4% incidence of serious granulocytopenia and rare cases of thrombocytopenic purpura.

BENEFIT OF CLOPIDOGREL

Although prasugrel is the focus of this review, the trials of prasugrel all compared its efficacy with that of clopidogrel. Furthermore, many patients should still receive clopidogrel and not prasugrel, so it is important to be familiar with the evidence of clopidogrel’s benefit.

Once approved for clinical use, clopidogrel was substituted for ticlopidine in patients undergoing coronary stenting on the basis of studies showing it to be at least as effective as ticlopidine and more tolerable. A series of trials of clopidogrel were done in patients across a spectrum of risk groups, from those at high risk of coronary heart disease to those presenting with ST-elevation MI. The time of pretreatment in the studies ranged from 3 hours to 6 days before percutaneous coronary intervention, and the duration of treatment following intervention ranged from 30 days to 1 year.

Clopidogrel in non-ST-elevation acute coronary syndromes

The CURE trial² (Clopidogrel in Unstable Angina to Prevent Recurrent Events), published in 2001, established clopidogrel as a therapy for unstable ischemic syndromes, whether treated medically or with revascularization. In that trial, 12,562 patients with acute coronary syndromes without ST elevation (ie, unstable angina or non-ST-elevation MI), as defined by electrocardiographic changes or positive cardiac markers, were randomized to receive clopidogrel (a 300-mg loading dose followed by 75-mg maintenance doses) or placebo for a mean duration of 9 months. All patients also received aspirin 75 mg to 325 mg daily.

The composite outcome of death from cardiovascular causes, nonfatal MI, or stroke occurred in 20% fewer patients treated with clopidogrel than with placebo (9.3% vs 11.4%). The benefit was similar in patients undergoing revascularization compared with those treated medically.

Although there were significantly more cases of major bleeding in the clopidogrel group than in the placebo group (3.7% vs 2.7%), the number of episodes of life-threatening bleeding or hemorrhagic strokes was the same.

PCI-CURE⁵ was a substudy of the CURE trial in patients who underwent a percutaneous coronary intervention. Patients were pretreated with clopidogrel or placebo for a mean of 6 days before the procedure. Afterward, they all received clopidogrel plus aspirin in an unblinded fashion for 2 to 4 weeks, and then the randomized study drug was resumed for a mean of 8 months.

Significantly fewer adverse events occurred in the clopidogrel group as tallied at the time of the intervention, 1 month later, and 8 months later.

 

 

Clopidogrel in ST-elevation acute MI

The CLARITY-TIMI 28 trial⁶ (Clopidogrel as Adjunctive Reperfusion Therapy—Thrombolysis in Myocardial Infarction 28) showed that adding clopidogrel (a 300-mg loading dose, then 75 mg daily) to aspirin benefitted patients with ST-elevation MI receiving fibrinolytic therapy. At 30 days, cardiovascular death, recurrent MI, or urgent revascularization had occurred in 11.6% of the clopidogrel group vs 14.1% of the placebo group, a statistically significant difference. The rates of major or minor bleeding were no higher in the clopidogrel group than in the placebo group, an especially remarkable finding in patients receiving thrombolytic therapy.

PCI-CLARITY.⁷ About half of the patients in the CLARITY trial ultimately underwent a percutaneous coronary intervention after fibrinolytic therapy, with results reported as the PCI-CLARITY substudy. Like those in PCI-CURE, these patients were randomized to receive pretreatment with either clopidogrel or placebo before the procedure, in this study for a median of 3 days. Both groups received clopidogrel afterward. At 30 days from randomization, the outcome of cardiovascular death, MI, or stroke had occurred in 7.5% of the clopidogrel group compared with 12.0% of the placebo group, which was statistically significant, without any significant excess in the rates of major or minor bleeding.

COMMIT⁸ (the Clopidogrel and Metoprolol in Myocardial Infarction Trial) also showed clopidogrel to be beneficial in patients with acute MI. This trial included more than 45,000 patients in China with acute MI, 93% of whom had ST-segment elevation. In contrast to CLARITY, in COMMIT barely more than half of the patients received fibrinolysis, fewer than 5% proceeded to percutaneous interventions, and no loading dose was given: patients in the clopidogrel group received 75 mg/day from the outset.

At 15 days, the incidence of death, reinfarction, or stroke was 9.2% with clopidogrel compared with 10.1% with placebo, a small but statistically significant difference. Again, the rate of major bleeding was not significantly higher, either overall or in patients over age 70.

Of note, patients over age 75 were excluded from CLARITY, and as mentioned, no loading dose was used in COMMIT. Thus, for patients receiving fibrinolysis who are over age 75, there is no evidence to support the safety of a loading dose, and clopidogrel should be started at 75 mg daily.

Clopidogrel in elective percutaneous coronary intervention

The CREDO trial⁹ (Clopidogrel for the Reduction of Events During Observation) was in patients referred for elective percutaneous coronary intervention. Three to 24 hours before the procedure, the patients received either a 300-mg loading dose of clopidogrel or placebo; afterward, all patients received clopidogrel 75 mg/day for 28 days. All patients also received aspirin.

A clopidogrel loading dose 3 to 24 hours before the intervention did not produce a statistically significant reduction in ischemic events, although a post hoc subgroup analysis suggested that patients who received the loading dose between 6 and 24 hours before did benefit, with a relative risk reduction of 38.6% in the composite end point (P = .051).

After 28 days, the patients who had received the clopidogrel loading dose were continued on clopidogrel, while those in the placebo group were switched back to placebo. At 1 year, the investigators found a significantly lower rate of the composite end point with the prolonged course of clopidogrel (8.5% vs 11.5%).

In summary, these studies found clopidogrel to be beneficial in a broad spectrum of coronary diseases. Subgroup analyses suggest that pretreatment before percutaneous coronary intervention provides additional benefit, particularly if clopidogrel is given at least 6 hours in advance (the time necessary for clopidogrel to cause substantial platelet inhibition).

SOME PATIENTS RESPOND LESS TO CLOPIDOGREL

The level of platelet inhibition induced by clopidogrel varies. In different studies, the frequency of clopidogrel “nonresponsiveness” ranged from 5% to 56% of patients, depending on which test and which cutoff values were used. The distribution of responses to clopidogrel is wide and fits a normal gaussian curve.¹⁰

A large fraction of the population carries a gene that may account for some of the interpatient variation in platelet inhibition with clopidogrel. Carriers of a reduced-function CYP2C19 allele—approximately 30% of people in one study—have significantly lower levels of the active metabolite of clopidogrel, less platelet inhibition from clopidogrel therapy, and a 53% higher rate of death from cardiovascular causes, MI, or stroke.¹¹

 

PRASUGREL, THE NEWEST THIENOPYRIDINE

Prasugrel, FDA-approved in July 2009 for the treatment of acute coronary syndromes, is given in an oral loading dose of 60 mg followed by an oral maintenance dose of 10 mg daily.

Pharmacology of prasugrel vs clopidogrel

As noted previously, the thienopyridines are prodrugs that require hepatic conversion to exert antiplatelet effects.

Metabolism. Prasugrel’s hepatic activation involves a single step, in contrast to the multiple-step process required for activation of clopidogrel. Clopidogrel is primarily hydrolyzed by intestinal and plasma esterases to an inactive terminal metabolite, with the residual unhydrolized drug undergoing a two-step metabolism that depends on cytochrome P450 enzymes. Prasugrel is also extensively hydrolyzed by these esterases, but the intermediate product is then metabolized in a single step to the active sulfhydryl compound, mainly by CYP3A4 and CYP2B6.

Thus, about 80% of an orally absorbed dose of prasugrel is converted to active drug, compared with only 10% to 20% of absorbed clopidogrel.

Time to peak effect. With clopidogrel, maximal inhibition of platelet aggregation occurs 3 to 5 days after starting therapy with 75 mg daily without a loading dose, but within 4 to 6 hours if a loading dose of 300 to 600 mg is given. In contrast, a prasugrel loading dose produces more than 80% of its platelet inhibitory effects by 30 minutes, and peak activity is observed within 4 hours.¹² The platelet inhibition induced by prasugrel at 30 minutes after administration is comparable to the peak effect of clopidogrel at 6 hours.¹³

Dose-response. Prasugrel’s inhibition of platelet aggregation is dose-related.

Prasugrel is about 10 times more potent than clopidogrel and 100 times more potent than ticlopidine. Thus, treatment with 5 mg of prasugrel results in inhibition of platelet activity (distributed in a gaussian curve) very similar to that produced by 75 mg of clopidogrel. On the other hand, even a maintenance dose of 150 mg of clopidogrel inhibits platelet activity to a lesser degree than 10 mg of prasugrel (46% vs 61%),¹⁴ so clopidogrel appears to reach a plateau of platelet inhibition that prasugrel can overcome.

At the approved dose of prasugrel, inhibition of platelet aggregation is significantly greater and there are fewer “nonresponders” than with clopidogrel.

Interactions. Drugs that inhibit CYP3A4 do not inhibit the efficacy of prasugrel, but they can inhibit that of clopidogrel. Some commonly used drugs that have this effect are the statins (eg, atorvastain [Lipitor]) and the macrolide antibiotics (eg, erythromycin). Furthermore, whereas proton pump inhibitors have been shown to diminish the effect of clopidogrel by reducing the formation of its active metabolite, no such effect has been noted with prasugrel.

Prasugrel in phase 2 trials: Finding the optimal dosage

A phase 2 trial compared three prasugrel regimens (loading dose/daily maintenance dose of 40 mg/7.5 mg, 60 mg/10 mg, and 60 mg/15 mg) and standard clopidogrel therapy (300 mg/75 mg) in patients undergoing elective or urgent percutaneous coronary intervention.¹⁵ No significant difference in outcomes was seen in the groups receiving the three prasugrel regimens. However, more “minimal bleeding events” (defined by the criteria of the TIMI trial¹⁶) occurred with high-dose prasugrel than with lower-dose prasugrel or with clopidogrel, leading to use of the intermediate-dose prasugrel regimen (60-mg loading dose, 10-mg daily maintenance) for later trials.

Another phase 2 trial randomized 201 patients undergoing elective percutaneous coronary intervention to receive prasugrel 60 mg/10 mg or clopidogrel 600 mg/150 mg.¹⁴ In all patients, the loading dose was given about 1 hour before cardiac catheterization. As soon as 30 minutes after the loading dose, platelet inhibition was superior with prasugrel (31% vs 5% inhibition of platelet aggregation), and it remained significantly higher at 6 hours (75% vs 32%) and during the maintenance phase (61% vs 46%).

 

 

Phase 3 trial of prasugrel vs clopidogrel: TRITON-TIMI 38

Only one large phase 3 trial of prasugrel has been completed: TRITON-TIMI 38 (the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel—Thrombolysis in Myocardial Infarction),¹⁷ which enrolled adults with moderate-risk to high-risk acute coronary syndromes scheduled to undergo a percutaneous coronary intervention. In this trial, 10,074 patients were enrolled who had moderate-to high-risk unstable angina or non-ST-elevation MI, and 3,534 patients were enrolled who had ST-elevation MI.

Patients were randomized to receive prasugrel (a 60-mg loading dose, then 10 mg daily) or clopidogrel (a 300-mg loading dose, then 75 mg daily) and were treated for 6 to 15 months. All patients also received aspirin.

The primary end point, a composite of death from cardiovascular causes, nonfatal MI, or nonfatal stroke, occurred in significantly fewer patients treated with prasugrel than with clopidogrel (9.9% vs 12.1%, P < .001) (Table 1). Most of the benefit was due to fewer nonfatal MIs during the follow-up period (7.4% vs 9.7%, P < .001). Additionally, the prasugrel group had a significantly lower rate of stent thrombosis compared with the clopidogrel group (1.1% vs 2.4%; P < .001).

These benefits came at a price of more bleeding. Of those patients who did not undergo coronary artery bypass grafting, more experienced bleeding in the prasugrel group than in the clopidogrel group (2.4% vs 1.8%, P = .03), including a higher rate of life-threatening bleeding (1.4% vs 0.89%, P = .01) and fatal bleeding (0.4% vs 0.1%, P = .002). More patients discontinued prasugrel because of hemorrhage (2.5% vs 1.4%, P < .001). In patients who proceeded to coronary artery bypass grafting, the rate of major bleeding was more than four times higher in those who received prasugrel than in those who received clopidogrel (13.4% vs 3.2%, P < .001).

A higher rate of adverse events related to colon cancer was also noted in patients treated with prasugrel, although the authors suggest this may have resulted from the stronger antiplatelet effects of prasugrel bringing more tumors to medical attention due to bleeding.

Overall death rates did not differ significantly between the treatment groups.

In a post hoc analysis,¹⁸ prasugrel was superior to clopidogrel in preventing ischemic events both during the first 3 days following randomization (the “loading phase”) and for the remainder of the trial (the “maintenance phase”). Whereas bleeding risk was similar with the two drugs during the loading phase, prasugrel was subsequently associated with more bleeding during the maintenance phase.

Certain patient subgroups had no net benefit or even suffered harm from prasugrel compared with clopidogrel.¹⁷ Patients with previous stroke or transient ischemic attack had net harm from prasugrel (hazard ratio 1.54, P = .04) and showed a strong trend toward a greater rate of major bleeding (P = .06). Patients age 75 and older and those weighing less than 60 kg had no net benefit from prasugrel.

Cost of prasugrel

Prasugrel is currently priced at 18% more than clopidogrel, with average wholesale prices per pill of $6.65 for prasugrel 10 mg compared with $5.63 for clopidogrel 75 mg. (Prasugrel 10-mg pills cost $6.33 at drugstore.com or $7.60 at CVS; clopidogrel 75-mg pills cost $5.33 at drugstore.com or $6.43 at CVS.) The patent on clopidogrel expires in November 2011, after which the price differential is expected to become significantly greater.

TICAGRELOR, A REVERSIBLE ORAL AGENT

Ticagrelor, the first reversible oral P2Y12 receptor antagonist, is an alternative to thienopyridine therapy for acute coronary syndromes.

Ticagrelor is quickly absorbed, does not require metabolic activation, and has a rapid antiplatelet effect and offset of effect, which closely follow drug-exposure levels. In a large randomized controlled trial in patients with acute coronary syndromes with or without STsegment elevation, treatment with ticagrelor compared with clopidogrel resulted in a significant reduction in death from vascular causes, MI, or stroke (9.8% vs 11.7%).¹⁹

Given its reversible effect on platelet inhibition, ticagrelor may be preferred in patients whose coronary anatomy is unknown and for whom coronary artery bypass grafting is deemed probable. It is still undergoing trials and is not yet approved.

 

TAKE-HOME POINTS

Prasugrel is more potent, more rapid in onset, and more consistent in inhibiting platelet aggregation than clopidogrel. A large clinical trial¹⁷ found prasugrel to be superior to clopidogrel for patients with moderate-to high-risk acute coronary syndromes with high probability of undergoing a percutaneous coronary intervention.

Who should receive prasugrel, and how?

Prasugrel should be given after angiography to patients with non-ST-elevation acute coronary syndromes or at presentation to patients with ST-elevation MI. When used for planned percutaneous coronary intervention, prasugrel should be given at least 30 minutes before the intervention, as was done in phase 2 trials (although its routine use in this situation is not recommended—see below).

It is given in a one-time loading dose of 60 mg by mouth and then maintained with 10 mg by mouth once daily for at least 1 year. (At least 9 months of treatment with a thienopyridine is indicated for patients with acute coronary syndromes who are medically treated, and at least 1 year is indicated following urgent or elective percutaneous coronary intervention, including balloon angioplasty and placement of a bare-metal or drug-eluting stent.)

Who should not receive prasugrel?

For now, prasugrel should be avoided in favor of clopidogrel in patients at higher risk of bleeding. It is clearly contraindicated in patients with prior transient ischemic attack or stroke, for whom the risk of serious bleeding seems to be prohibitive. It should generally be avoided in patients age 75 and older, although it might be considered in those at particularly high risk of stent thrombosis, such as those with diabetes or prior MI. In patients weighing less than 60 kg, the package insert advises a reduced dose (5 mg), although clinical evidence for this practice is lacking.

As yet, we have no data assuring that prasugrel is safe to use in combination with fibrinolytic agents, so patients on thrombolytic therapy for acute MI should continue to receive clopidogrel starting immediately after lysis. Furthermore, in patients who proceeded to coronary artery bypass grafting, the rate of major bleeding was more than four times higher in the prasugrel group than in the clopidogrel group in the TRITON-TIMI 38 trial.¹⁷ No thienopyridine should be given to patients likely to proceed to coronary artery bypass grafting.

Only clopidogrel has evidence supporting its use as an alternative to aspirin for patients with atherosclerotic disease who cannot tolerate aspirin. Neither drug has evidence for use for primary prevention.

Other areas of uncertainty

Prior to angiography. Indications for prasugrel are currently limited by the narrow scope of the trial data. TRITON-TIMI 38,¹⁷ the only large trial completed to date, randomized patients to receive prasugrel only after their coronary anatomy was known, except for ST-elevation MI patients. It is unknown whether the benefits of prasugrel will outweigh the higher risk of bleeding in patients with acute coronary syndromes who do not proceed to percutaneous coronary interventions.

A clinical trial is currently under way comparing prasugrel with clopidogrel in 10,000 patients with acute coronary syndromes who will be medically managed without planned revascularization: A Comparison of Prasugrel and Clopidogrel in Acute Coronary Syndrome Subjects (TRILOGY ACS), ClinicalTrials.gov Identifier: NCT00699998. The trial has an estimated completion date of March 2011.

In cases of non-ST-elevation acute coronary syndrome, it is reasonable to wait to give a thienopyridine until after the coronary anatomy has been defined, if angiography will be completed soon after presentation. For example, a 1-hour delay before giving prasugrel still delivers antiplatelet therapy more quickly than giving clopidogrel on presentation. If longer delays are expected before angiography, however, the patient should be given a loading dose of clopidogrel “up front,” in accordance with guidelines published by the American College of Cardiology, American Heart Association, and European Society of Cardiology,²⁰ which recommend starting a thienopyridine early during hospitalization based on trial data with clopidogrel.

Patients undergoing elective percutaneous coronary intervention are at lower risk of stent thrombosis and other ischemic complications, so it is possible that the benefits of prasugrel would not outweigh the risks in these patients. Thus, prasugrel cannot yet be recommended for routine elective percutaneous coronary intervention except in individual cases in which the interventionalist feels that the patient may be at higher risk of thrombosis.

Prasugrel (Effient) is more potent and consistent in its effects than clopidogrel (Plavix), thus preventing more thrombotic events—but at a price of more bleeding. Therefore, the drugs must be appropriately selected for the individual patient.

Over the last 9 years, the thienopyridines—ticlopidine (Ticlid), clopidogrel, and now prasugrel—have become essential tools for treating acute coronary syndromes.

The usual underlying mechanism of acute coronary syndromes is thrombosis, caused by rupture of atherosclerotic plaque.¹ Accordingly, antithrombotic agents—aspirin, heparin, lowmolecular-weight heparin, glycoprotein IIb/IIIa inhibitors, the direct thrombin inhibitor bivalirudin (Angiomax), and thienopyridines—have all been shown to reduce the risk of major adverse cardiac outcomes in this setting.

In this article, we review the pharmacology and evidence of effectiveness of the thienopyridine drugs, focusing on prasugrel, the latest thienopyridine to be approved by the US Food and Drug Administration (FDA).

THIENOPYRIDINES INHIBIT PLATELET ACTIVATION AND AGGREGATION

Thienopyridines are prodrugs that require conversion by hepatic cytochrome P450 enzymes. The active metabolites bind irreversibly to platelet P2Y12 receptors. Consequently, they permanently block signalling mediated by platelet adenosine diphosphate-P2Y12 receptors, thereby inhibiting glycoprotein IIb/IIIa receptor activation and platelet aggregation.

Aspirin, in contrast, inhibits platelets by blocking the thromboxane-mediated pathway. Therefore, the combination of aspirin plus a thienopyridine has an additive effect.²

The effect of thienopyridines on platelets is irreversible. Therefore, although the half-life of prasugrel’s active metabolite is 3.7 hours, its inhibitory effects last for 96 hours, essentially the time for half the body’s circulating platelets to be replaced.

TICLOPIDINE, THE FIRST THIENOPYRIDINE

Ticlopidine was the first thienopyridine to be approved by the FDA. Its initial studies in unstable angina were small, their designs did not call for patients to concurrently receive aspirin, and all they showed was that ticlopidine was about as beneficial as aspirin. Consequently, the studies had little impact on clinical practice.³

In a pivotal trial,⁴ patients who received coronary stents were randomized to afterward receive either the combination of ticlopidine plus aspirin or anticoagulation therapy with heparin, phenprocoumon (a coumarin derivative available in Europe), and aspirin. At 30 days, an ischemic complication (death, myocardial infarction [MI], repeat intervention) had occurred in 6.2% of the anticoagulation therapy group vs 1.6% of the ticlopidine group, a risk reduction of 75%. Rates of stent occlusion, MI, and revascularization were 80% to 85% lower in the ticlodipine group. This study paved the way for widespread use of thienopyridines.

Ticlopidine’s use was limited, however, by a 2.4% incidence of serious granulocytopenia and rare cases of thrombocytopenic purpura.

BENEFIT OF CLOPIDOGREL

Although prasugrel is the focus of this review, the trials of prasugrel all compared its efficacy with that of clopidogrel. Furthermore, many patients should still receive clopidogrel and not prasugrel, so it is important to be familiar with the evidence of clopidogrel’s benefit.

Once approved for clinical use, clopidogrel was substituted for ticlopidine in patients undergoing coronary stenting on the basis of studies showing it to be at least as effective as ticlopidine and more tolerable. A series of trials of clopidogrel were done in patients across a spectrum of risk groups, from those at high risk of coronary heart disease to those presenting with ST-elevation MI. The time of pretreatment in the studies ranged from 3 hours to 6 days before percutaneous coronary intervention, and the duration of treatment following intervention ranged from 30 days to 1 year.

Clopidogrel in non-ST-elevation acute coronary syndromes

The CURE trial² (Clopidogrel in Unstable Angina to Prevent Recurrent Events), published in 2001, established clopidogrel as a therapy for unstable ischemic syndromes, whether treated medically or with revascularization. In that trial, 12,562 patients with acute coronary syndromes without ST elevation (ie, unstable angina or non-ST-elevation MI), as defined by electrocardiographic changes or positive cardiac markers, were randomized to receive clopidogrel (a 300-mg loading dose followed by 75-mg maintenance doses) or placebo for a mean duration of 9 months. All patients also received aspirin 75 mg to 325 mg daily.

The composite outcome of death from cardiovascular causes, nonfatal MI, or stroke occurred in 20% fewer patients treated with clopidogrel than with placebo (9.3% vs 11.4%). The benefit was similar in patients undergoing revascularization compared with those treated medically.

Although there were significantly more cases of major bleeding in the clopidogrel group than in the placebo group (3.7% vs 2.7%), the number of episodes of life-threatening bleeding or hemorrhagic strokes was the same.

PCI-CURE⁵ was a substudy of the CURE trial in patients who underwent a percutaneous coronary intervention. Patients were pretreated with clopidogrel or placebo for a mean of 6 days before the procedure. Afterward, they all received clopidogrel plus aspirin in an unblinded fashion for 2 to 4 weeks, and then the randomized study drug was resumed for a mean of 8 months.

Significantly fewer adverse events occurred in the clopidogrel group as tallied at the time of the intervention, 1 month later, and 8 months later.

 

 

Clopidogrel in ST-elevation acute MI

The CLARITY-TIMI 28 trial⁶ (Clopidogrel as Adjunctive Reperfusion Therapy—Thrombolysis in Myocardial Infarction 28) showed that adding clopidogrel (a 300-mg loading dose, then 75 mg daily) to aspirin benefitted patients with ST-elevation MI receiving fibrinolytic therapy. At 30 days, cardiovascular death, recurrent MI, or urgent revascularization had occurred in 11.6% of the clopidogrel group vs 14.1% of the placebo group, a statistically significant difference. The rates of major or minor bleeding were no higher in the clopidogrel group than in the placebo group, an especially remarkable finding in patients receiving thrombolytic therapy.

PCI-CLARITY.⁷ About half of the patients in the CLARITY trial ultimately underwent a percutaneous coronary intervention after fibrinolytic therapy, with results reported as the PCI-CLARITY substudy. Like those in PCI-CURE, these patients were randomized to receive pretreatment with either clopidogrel or placebo before the procedure, in this study for a median of 3 days. Both groups received clopidogrel afterward. At 30 days from randomization, the outcome of cardiovascular death, MI, or stroke had occurred in 7.5% of the clopidogrel group compared with 12.0% of the placebo group, which was statistically significant, without any significant excess in the rates of major or minor bleeding.

COMMIT⁸ (the Clopidogrel and Metoprolol in Myocardial Infarction Trial) also showed clopidogrel to be beneficial in patients with acute MI. This trial included more than 45,000 patients in China with acute MI, 93% of whom had ST-segment elevation. In contrast to CLARITY, in COMMIT barely more than half of the patients received fibrinolysis, fewer than 5% proceeded to percutaneous interventions, and no loading dose was given: patients in the clopidogrel group received 75 mg/day from the outset.

At 15 days, the incidence of death, reinfarction, or stroke was 9.2% with clopidogrel compared with 10.1% with placebo, a small but statistically significant difference. Again, the rate of major bleeding was not significantly higher, either overall or in patients over age 70.

Of note, patients over age 75 were excluded from CLARITY, and as mentioned, no loading dose was used in COMMIT. Thus, for patients receiving fibrinolysis who are over age 75, there is no evidence to support the safety of a loading dose, and clopidogrel should be started at 75 mg daily.

Clopidogrel in elective percutaneous coronary intervention

The CREDO trial⁹ (Clopidogrel for the Reduction of Events During Observation) was in patients referred for elective percutaneous coronary intervention. Three to 24 hours before the procedure, the patients received either a 300-mg loading dose of clopidogrel or placebo; afterward, all patients received clopidogrel 75 mg/day for 28 days. All patients also received aspirin.

A clopidogrel loading dose 3 to 24 hours before the intervention did not produce a statistically significant reduction in ischemic events, although a post hoc subgroup analysis suggested that patients who received the loading dose between 6 and 24 hours before did benefit, with a relative risk reduction of 38.6% in the composite end point (P = .051).

After 28 days, the patients who had received the clopidogrel loading dose were continued on clopidogrel, while those in the placebo group were switched back to placebo. At 1 year, the investigators found a significantly lower rate of the composite end point with the prolonged course of clopidogrel (8.5% vs 11.5%).

In summary, these studies found clopidogrel to be beneficial in a broad spectrum of coronary diseases. Subgroup analyses suggest that pretreatment before percutaneous coronary intervention provides additional benefit, particularly if clopidogrel is given at least 6 hours in advance (the time necessary for clopidogrel to cause substantial platelet inhibition).

SOME PATIENTS RESPOND LESS TO CLOPIDOGREL

The level of platelet inhibition induced by clopidogrel varies. In different studies, the frequency of clopidogrel “nonresponsiveness” ranged from 5% to 56% of patients, depending on which test and which cutoff values were used. The distribution of responses to clopidogrel is wide and fits a normal gaussian curve.¹⁰

A large fraction of the population carries a gene that may account for some of the interpatient variation in platelet inhibition with clopidogrel. Carriers of a reduced-function CYP2C19 allele—approximately 30% of people in one study—have significantly lower levels of the active metabolite of clopidogrel, less platelet inhibition from clopidogrel therapy, and a 53% higher rate of death from cardiovascular causes, MI, or stroke.¹¹

 

PRASUGREL, THE NEWEST THIENOPYRIDINE

Prasugrel, FDA-approved in July 2009 for the treatment of acute coronary syndromes, is given in an oral loading dose of 60 mg followed by an oral maintenance dose of 10 mg daily.

Pharmacology of prasugrel vs clopidogrel

As noted previously, the thienopyridines are prodrugs that require hepatic conversion to exert antiplatelet effects.

Metabolism. Prasugrel’s hepatic activation involves a single step, in contrast to the multiple-step process required for activation of clopidogrel. Clopidogrel is primarily hydrolyzed by intestinal and plasma esterases to an inactive terminal metabolite, with the residual unhydrolized drug undergoing a two-step metabolism that depends on cytochrome P450 enzymes. Prasugrel is also extensively hydrolyzed by these esterases, but the intermediate product is then metabolized in a single step to the active sulfhydryl compound, mainly by CYP3A4 and CYP2B6.

Thus, about 80% of an orally absorbed dose of prasugrel is converted to active drug, compared with only 10% to 20% of absorbed clopidogrel.

Time to peak effect. With clopidogrel, maximal inhibition of platelet aggregation occurs 3 to 5 days after starting therapy with 75 mg daily without a loading dose, but within 4 to 6 hours if a loading dose of 300 to 600 mg is given. In contrast, a prasugrel loading dose produces more than 80% of its platelet inhibitory effects by 30 minutes, and peak activity is observed within 4 hours.¹² The platelet inhibition induced by prasugrel at 30 minutes after administration is comparable to the peak effect of clopidogrel at 6 hours.¹³

Dose-response. Prasugrel’s inhibition of platelet aggregation is dose-related.

Prasugrel is about 10 times more potent than clopidogrel and 100 times more potent than ticlopidine. Thus, treatment with 5 mg of prasugrel results in inhibition of platelet activity (distributed in a gaussian curve) very similar to that produced by 75 mg of clopidogrel. On the other hand, even a maintenance dose of 150 mg of clopidogrel inhibits platelet activity to a lesser degree than 10 mg of prasugrel (46% vs 61%),¹⁴ so clopidogrel appears to reach a plateau of platelet inhibition that prasugrel can overcome.

At the approved dose of prasugrel, inhibition of platelet aggregation is significantly greater and there are fewer “nonresponders” than with clopidogrel.

Interactions. Drugs that inhibit CYP3A4 do not inhibit the efficacy of prasugrel, but they can inhibit that of clopidogrel. Some commonly used drugs that have this effect are the statins (eg, atorvastain [Lipitor]) and the macrolide antibiotics (eg, erythromycin). Furthermore, whereas proton pump inhibitors have been shown to diminish the effect of clopidogrel by reducing the formation of its active metabolite, no such effect has been noted with prasugrel.

Prasugrel in phase 2 trials: Finding the optimal dosage

A phase 2 trial compared three prasugrel regimens (loading dose/daily maintenance dose of 40 mg/7.5 mg, 60 mg/10 mg, and 60 mg/15 mg) and standard clopidogrel therapy (300 mg/75 mg) in patients undergoing elective or urgent percutaneous coronary intervention.¹⁵ No significant difference in outcomes was seen in the groups receiving the three prasugrel regimens. However, more “minimal bleeding events” (defined by the criteria of the TIMI trial¹⁶) occurred with high-dose prasugrel than with lower-dose prasugrel or with clopidogrel, leading to use of the intermediate-dose prasugrel regimen (60-mg loading dose, 10-mg daily maintenance) for later trials.

Another phase 2 trial randomized 201 patients undergoing elective percutaneous coronary intervention to receive prasugrel 60 mg/10 mg or clopidogrel 600 mg/150 mg.¹⁴ In all patients, the loading dose was given about 1 hour before cardiac catheterization. As soon as 30 minutes after the loading dose, platelet inhibition was superior with prasugrel (31% vs 5% inhibition of platelet aggregation), and it remained significantly higher at 6 hours (75% vs 32%) and during the maintenance phase (61% vs 46%).

 

 

Phase 3 trial of prasugrel vs clopidogrel: TRITON-TIMI 38

Only one large phase 3 trial of prasugrel has been completed: TRITON-TIMI 38 (the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel—Thrombolysis in Myocardial Infarction),¹⁷ which enrolled adults with moderate-risk to high-risk acute coronary syndromes scheduled to undergo a percutaneous coronary intervention. In this trial, 10,074 patients were enrolled who had moderate-to high-risk unstable angina or non-ST-elevation MI, and 3,534 patients were enrolled who had ST-elevation MI.

Patients were randomized to receive prasugrel (a 60-mg loading dose, then 10 mg daily) or clopidogrel (a 300-mg loading dose, then 75 mg daily) and were treated for 6 to 15 months. All patients also received aspirin.

The primary end point, a composite of death from cardiovascular causes, nonfatal MI, or nonfatal stroke, occurred in significantly fewer patients treated with prasugrel than with clopidogrel (9.9% vs 12.1%, P < .001) (Table 1). Most of the benefit was due to fewer nonfatal MIs during the follow-up period (7.4% vs 9.7%, P < .001). Additionally, the prasugrel group had a significantly lower rate of stent thrombosis compared with the clopidogrel group (1.1% vs 2.4%; P < .001).

These benefits came at a price of more bleeding. Of those patients who did not undergo coronary artery bypass grafting, more experienced bleeding in the prasugrel group than in the clopidogrel group (2.4% vs 1.8%, P = .03), including a higher rate of life-threatening bleeding (1.4% vs 0.89%, P = .01) and fatal bleeding (0.4% vs 0.1%, P = .002). More patients discontinued prasugrel because of hemorrhage (2.5% vs 1.4%, P < .001). In patients who proceeded to coronary artery bypass grafting, the rate of major bleeding was more than four times higher in those who received prasugrel than in those who received clopidogrel (13.4% vs 3.2%, P < .001).

A higher rate of adverse events related to colon cancer was also noted in patients treated with prasugrel, although the authors suggest this may have resulted from the stronger antiplatelet effects of prasugrel bringing more tumors to medical attention due to bleeding.

Overall death rates did not differ significantly between the treatment groups.

In a post hoc analysis,¹⁸ prasugrel was superior to clopidogrel in preventing ischemic events both during the first 3 days following randomization (the “loading phase”) and for the remainder of the trial (the “maintenance phase”). Whereas bleeding risk was similar with the two drugs during the loading phase, prasugrel was subsequently associated with more bleeding during the maintenance phase.

Certain patient subgroups had no net benefit or even suffered harm from prasugrel compared with clopidogrel.¹⁷ Patients with previous stroke or transient ischemic attack had net harm from prasugrel (hazard ratio 1.54, P = .04) and showed a strong trend toward a greater rate of major bleeding (P = .06). Patients age 75 and older and those weighing less than 60 kg had no net benefit from prasugrel.

Cost of prasugrel

Prasugrel is currently priced at 18% more than clopidogrel, with average wholesale prices per pill of $6.65 for prasugrel 10 mg compared with $5.63 for clopidogrel 75 mg. (Prasugrel 10-mg pills cost $6.33 at drugstore.com or $7.60 at CVS; clopidogrel 75-mg pills cost $5.33 at drugstore.com or $6.43 at CVS.) The patent on clopidogrel expires in November 2011, after which the price differential is expected to become significantly greater.

TICAGRELOR, A REVERSIBLE ORAL AGENT

Ticagrelor, the first reversible oral P2Y12 receptor antagonist, is an alternative to thienopyridine therapy for acute coronary syndromes.

Ticagrelor is quickly absorbed, does not require metabolic activation, and has a rapid antiplatelet effect and offset of effect, which closely follow drug-exposure levels. In a large randomized controlled trial in patients with acute coronary syndromes with or without STsegment elevation, treatment with ticagrelor compared with clopidogrel resulted in a significant reduction in death from vascular causes, MI, or stroke (9.8% vs 11.7%).¹⁹

Given its reversible effect on platelet inhibition, ticagrelor may be preferred in patients whose coronary anatomy is unknown and for whom coronary artery bypass grafting is deemed probable. It is still undergoing trials and is not yet approved.

 

TAKE-HOME POINTS

Prasugrel is more potent, more rapid in onset, and more consistent in inhibiting platelet aggregation than clopidogrel. A large clinical trial¹⁷ found prasugrel to be superior to clopidogrel for patients with moderate-to high-risk acute coronary syndromes with high probability of undergoing a percutaneous coronary intervention.

Who should receive prasugrel, and how?

Prasugrel should be given after angiography to patients with non-ST-elevation acute coronary syndromes or at presentation to patients with ST-elevation MI. When used for planned percutaneous coronary intervention, prasugrel should be given at least 30 minutes before the intervention, as was done in phase 2 trials (although its routine use in this situation is not recommended—see below).

It is given in a one-time loading dose of 60 mg by mouth and then maintained with 10 mg by mouth once daily for at least 1 year. (At least 9 months of treatment with a thienopyridine is indicated for patients with acute coronary syndromes who are medically treated, and at least 1 year is indicated following urgent or elective percutaneous coronary intervention, including balloon angioplasty and placement of a bare-metal or drug-eluting stent.)

Who should not receive prasugrel?

For now, prasugrel should be avoided in favor of clopidogrel in patients at higher risk of bleeding. It is clearly contraindicated in patients with prior transient ischemic attack or stroke, for whom the risk of serious bleeding seems to be prohibitive. It should generally be avoided in patients age 75 and older, although it might be considered in those at particularly high risk of stent thrombosis, such as those with diabetes or prior MI. In patients weighing less than 60 kg, the package insert advises a reduced dose (5 mg), although clinical evidence for this practice is lacking.

As yet, we have no data assuring that prasugrel is safe to use in combination with fibrinolytic agents, so patients on thrombolytic therapy for acute MI should continue to receive clopidogrel starting immediately after lysis. Furthermore, in patients who proceeded to coronary artery bypass grafting, the rate of major bleeding was more than four times higher in the prasugrel group than in the clopidogrel group in the TRITON-TIMI 38 trial.¹⁷ No thienopyridine should be given to patients likely to proceed to coronary artery bypass grafting.

Only clopidogrel has evidence supporting its use as an alternative to aspirin for patients with atherosclerotic disease who cannot tolerate aspirin. Neither drug has evidence for use for primary prevention.

Other areas of uncertainty

Prior to angiography. Indications for prasugrel are currently limited by the narrow scope of the trial data. TRITON-TIMI 38,¹⁷ the only large trial completed to date, randomized patients to receive prasugrel only after their coronary anatomy was known, except for ST-elevation MI patients. It is unknown whether the benefits of prasugrel will outweigh the higher risk of bleeding in patients with acute coronary syndromes who do not proceed to percutaneous coronary interventions.

A clinical trial is currently under way comparing prasugrel with clopidogrel in 10,000 patients with acute coronary syndromes who will be medically managed without planned revascularization: A Comparison of Prasugrel and Clopidogrel in Acute Coronary Syndrome Subjects (TRILOGY ACS), ClinicalTrials.gov Identifier: NCT00699998. The trial has an estimated completion date of March 2011.

In cases of non-ST-elevation acute coronary syndrome, it is reasonable to wait to give a thienopyridine until after the coronary anatomy has been defined, if angiography will be completed soon after presentation. For example, a 1-hour delay before giving prasugrel still delivers antiplatelet therapy more quickly than giving clopidogrel on presentation. If longer delays are expected before angiography, however, the patient should be given a loading dose of clopidogrel “up front,” in accordance with guidelines published by the American College of Cardiology, American Heart Association, and European Society of Cardiology,²⁰ which recommend starting a thienopyridine early during hospitalization based on trial data with clopidogrel.

Patients undergoing elective percutaneous coronary intervention are at lower risk of stent thrombosis and other ischemic complications, so it is possible that the benefits of prasugrel would not outweigh the risks in these patients. Thus, prasugrel cannot yet be recommended for routine elective percutaneous coronary intervention except in individual cases in which the interventionalist feels that the patient may be at higher risk of thrombosis.

Oligonucleotide microarray analysis reveals a higher yield of abnormalities in children with autism than previously seen with other techniques.

LOUISVILLE—Microarray analysis and testing for fragile X syndrome revealed genetic abnormalities in 10% of children with autism, according to research presented at the 38th National Meeting of the Child Neurology Society. Christa Lese Martin, PhD, and colleagues studied 93 children (79 males) with autism spectrum disorders (ASD) between ages 2 and 7 who were enrolled in an ongoing, comprehensive project aimed at finding meaningful subtypes of autism.

The children underwent a panel of genetic testing, including high-resolution karyotype and fluorescence in situ hybridization (FISH) for 15q, fragile X testing, and array comparative genomic hybridization using the EmArray Cyto array, a custom array of 44,000 oligos designed by Emory Genetics Laboratory in Atlanta, to detect copy number imbalances across the genome.

“Our study demonstrates that clinical microarray—ie, cytogenetic array—testing to look for genome-wide deletions or duplications and fragile X testing is strongly indicated in individuals with ASD,” Dr. Martin, Associate Professor of Human Genetics, Emory University School of Medicine in Atlanta, told Neurology Reviews. “The combination of these two tests identifies a cause for ASD in approximately 10% of individuals. The identification of a cause for ASD not only provides a clinical diagnosis, but also provides the opportunity for accurate genetic counseling for the families.”

New ASD Loci Identified
The pathogenic abnormalities identified included fragile X syndrome, a 17p deletion of the Smith-Magenis region, an unbalanced translocation resulting in duplication of 2p and deletion of 9p, and 16p11.2 duplication syndrome. The children with imbalances had similar cognitive and behavioral profiles as the remainder of the sample and had no obvious clues suggesting any underlying genetic abnormalities, the investigators noted.

“Karyotype alone is not sensitive, and multiple FISH probes would be needed to identify the most common autism loci, greatly increasing cost while still missing the important opportunity to examine the rest of the genome,” reported Dr. Martin’s group. “Our results also demonstrate that genome-wide microarray technology has increased yield and provides more cost-effective testing.”

In addition, four subjects had fragile X mutations; one was a full mutation, one was a premutation, and two were grey zone mutations. “Full mutation fragile X is a relatively rare finding in autism samples,” the researchers noted. “But premutation and grey zone mutations are more common than would be expected in a general population sample.”

Array Analysis in Diagnosing ASD
“Cytogenetic array analysis and fragile X testing are currently offered in many clinical genetic laboratories and should be offered as part of the diagnostic workup in the evaluation of ASD,” asserted Dr. Martin, who co-leads the array services at Emory Genetics Laboratory. “In addition, since all of the cases examined in our study had normal G-banded chromosome analysis, but several clinically significant abnormalities were identified by array analysis, these data provide further support that the array should be used as the first-line cytogenetic test since this analysis can identify imbalances that are below the resolution of a routine karyotype.

“The yield of cytogenetic array testing in individuals with ASD is quite high—8% to 10% has now been reported from various studies,” she added. “This information is invaluable to families to alleviate their search for a cause in their children and counsel them appropriately on recurrence risks in their family.”

Oligonucleotide microarray analysis reveals a higher yield of abnormalities in children with autism than previously seen with other techniques.

LOUISVILLE—Microarray analysis and testing for fragile X syndrome revealed genetic abnormalities in 10% of children with autism, according to research presented at the 38th National Meeting of the Child Neurology Society. Christa Lese Martin, PhD, and colleagues studied 93 children (79 males) with autism spectrum disorders (ASD) between ages 2 and 7 who were enrolled in an ongoing, comprehensive project aimed at finding meaningful subtypes of autism.

The children underwent a panel of genetic testing, including high-resolution karyotype and fluorescence in situ hybridization (FISH) for 15q, fragile X testing, and array comparative genomic hybridization using the EmArray Cyto array, a custom array of 44,000 oligos designed by Emory Genetics Laboratory in Atlanta, to detect copy number imbalances across the genome.

“Our study demonstrates that clinical microarray—ie, cytogenetic array—testing to look for genome-wide deletions or duplications and fragile X testing is strongly indicated in individuals with ASD,” Dr. Martin, Associate Professor of Human Genetics, Emory University School of Medicine in Atlanta, told Neurology Reviews. “The combination of these two tests identifies a cause for ASD in approximately 10% of individuals. The identification of a cause for ASD not only provides a clinical diagnosis, but also provides the opportunity for accurate genetic counseling for the families.”

New ASD Loci Identified
The pathogenic abnormalities identified included fragile X syndrome, a 17p deletion of the Smith-Magenis region, an unbalanced translocation resulting in duplication of 2p and deletion of 9p, and 16p11.2 duplication syndrome. The children with imbalances had similar cognitive and behavioral profiles as the remainder of the sample and had no obvious clues suggesting any underlying genetic abnormalities, the investigators noted.

“Karyotype alone is not sensitive, and multiple FISH probes would be needed to identify the most common autism loci, greatly increasing cost while still missing the important opportunity to examine the rest of the genome,” reported Dr. Martin’s group. “Our results also demonstrate that genome-wide microarray technology has increased yield and provides more cost-effective testing.”

In addition, four subjects had fragile X mutations; one was a full mutation, one was a premutation, and two were grey zone mutations. “Full mutation fragile X is a relatively rare finding in autism samples,” the researchers noted. “But premutation and grey zone mutations are more common than would be expected in a general population sample.”

Array Analysis in Diagnosing ASD
“Cytogenetic array analysis and fragile X testing are currently offered in many clinical genetic laboratories and should be offered as part of the diagnostic workup in the evaluation of ASD,” asserted Dr. Martin, who co-leads the array services at Emory Genetics Laboratory. “In addition, since all of the cases examined in our study had normal G-banded chromosome analysis, but several clinically significant abnormalities were identified by array analysis, these data provide further support that the array should be used as the first-line cytogenetic test since this analysis can identify imbalances that are below the resolution of a routine karyotype.

“The yield of cytogenetic array testing in individuals with ASD is quite high—8% to 10% has now been reported from various studies,” she added. “This information is invaluable to families to alleviate their search for a cause in their children and counsel them appropriately on recurrence risks in their family.”

Oligonucleotide microarray analysis reveals a higher yield of abnormalities in children with autism than previously seen with other techniques.

LOUISVILLE—Microarray analysis and testing for fragile X syndrome revealed genetic abnormalities in 10% of children with autism, according to research presented at the 38th National Meeting of the Child Neurology Society. Christa Lese Martin, PhD, and colleagues studied 93 children (79 males) with autism spectrum disorders (ASD) between ages 2 and 7 who were enrolled in an ongoing, comprehensive project aimed at finding meaningful subtypes of autism.

The children underwent a panel of genetic testing, including high-resolution karyotype and fluorescence in situ hybridization (FISH) for 15q, fragile X testing, and array comparative genomic hybridization using the EmArray Cyto array, a custom array of 44,000 oligos designed by Emory Genetics Laboratory in Atlanta, to detect copy number imbalances across the genome.

“Our study demonstrates that clinical microarray—ie, cytogenetic array—testing to look for genome-wide deletions or duplications and fragile X testing is strongly indicated in individuals with ASD,” Dr. Martin, Associate Professor of Human Genetics, Emory University School of Medicine in Atlanta, told Neurology Reviews. “The combination of these two tests identifies a cause for ASD in approximately 10% of individuals. The identification of a cause for ASD not only provides a clinical diagnosis, but also provides the opportunity for accurate genetic counseling for the families.”

New ASD Loci Identified
The pathogenic abnormalities identified included fragile X syndrome, a 17p deletion of the Smith-Magenis region, an unbalanced translocation resulting in duplication of 2p and deletion of 9p, and 16p11.2 duplication syndrome. The children with imbalances had similar cognitive and behavioral profiles as the remainder of the sample and had no obvious clues suggesting any underlying genetic abnormalities, the investigators noted.

“Karyotype alone is not sensitive, and multiple FISH probes would be needed to identify the most common autism loci, greatly increasing cost while still missing the important opportunity to examine the rest of the genome,” reported Dr. Martin’s group. “Our results also demonstrate that genome-wide microarray technology has increased yield and provides more cost-effective testing.”

In addition, four subjects had fragile X mutations; one was a full mutation, one was a premutation, and two were grey zone mutations. “Full mutation fragile X is a relatively rare finding in autism samples,” the researchers noted. “But premutation and grey zone mutations are more common than would be expected in a general population sample.”

Array Analysis in Diagnosing ASD
“Cytogenetic array analysis and fragile X testing are currently offered in many clinical genetic laboratories and should be offered as part of the diagnostic workup in the evaluation of ASD,” asserted Dr. Martin, who co-leads the array services at Emory Genetics Laboratory. “In addition, since all of the cases examined in our study had normal G-banded chromosome analysis, but several clinically significant abnormalities were identified by array analysis, these data provide further support that the array should be used as the first-line cytogenetic test since this analysis can identify imbalances that are below the resolution of a routine karyotype.

“The yield of cytogenetic array testing in individuals with ASD is quite high—8% to 10% has now been reported from various studies,” she added. “This information is invaluable to families to alleviate their search for a cause in their children and counsel them appropriately on recurrence risks in their family.”

LOUISVILLE—The autistic brain builds a stronger-than-normal association between motor commands and proprioceptive feedback, which may account for why children with autism have difficulty forming the models necessary to engage not only in motor behavior, but in social and communicative behaviors, according to research presented at the 38th National Meeting of the Child Neurology Society.

Stewart H. Mostofsky, MD, Associate Professor of Neurology at Kennedy Krieger Institute and the Johns Hopkins University School of Medicine in Baltimore, and colleagues, observed 14 children with autism spectrum disorders and 13 typically developing children as they learned to control a robotic arm. Subjects attempted to reach a target of interest while the robotic arm produced a force perpendicular to that location.

To test this hypothesis, children engaged in a second experiment in which Dr. Mostofsky’s group observed and assessed generalization, the signature of activation fields of neurons. “Generalization can tell you about how [children] learn, because you can look at the way they are able to transfer what they learn in one particular state to another,” Dr. Mostofsky said.

“The generalization patterns were strikingly different,” Dr. Mostofsky and colleagues reported. Typically developing children generalized in both proprioceptive and visual coordinates when generating models of behavior; whereas, children with autism spectrum disorders only generalized in proprioceptive coordinates, and approximately twice as strong as the typically developing children. Furthermore, the tendency to generalize in proprioceptive coordinates was highly predictive of autism-associated impairments in performance in skilled motor gestures to imitation, as well as performance of these gestures to command, and with actual tool use (often referred to as “dyspraxia”).

“[Moreover], notions of feed-forward hypotheses would suggest that these same internal models that are the basis of learning skilled movements might also be the basis for which our brain learns to understand and recognize the actions of others,” Dr. Mostofsky stated. Therefore, impaired acquisition of skill movements may contribute to the social deficits associated with autism.

Serum IL-6 Levels

Consistent with this hypothesis, generalization in intrinsic proprioceptive coordinates was highly predictive of higher (more impaired) Autism Diagnostic Observation Schedule scores for children with autism, and predictive of higher (more impaired) Social Responsiveness Scale scores for children with autism and in typically developing children, according to Dr. Mostofsky.

Dr. Mostofsky and colleagues are now examining whether these findings are specific to autism. In addition, they want to determine whether the formation of internal models of action are associated with abnormal patterns of neural connectivity. “Our preliminary diffusion tensor imaging findings do suggest that disorganization of white matter in the primary sensorimotor cortex may be associated with this increased proprioceptive bias,” Dr. Mostofsky commented.

The researchers also want to determine whether these observations can be used to modify the learning patterns in autism, either on a behavioral level, or as cortical stimulation methods used to upregulate visual-premotor connections.

—Laura Sassano

LOUISVILLE—The autistic brain builds a stronger-than-normal association between motor commands and proprioceptive feedback, which may account for why children with autism have difficulty forming the models necessary to engage not only in motor behavior, but in social and communicative behaviors, according to research presented at the 38th National Meeting of the Child Neurology Society.

Stewart H. Mostofsky, MD, Associate Professor of Neurology at Kennedy Krieger Institute and the Johns Hopkins University School of Medicine in Baltimore, and colleagues, observed 14 children with autism spectrum disorders and 13 typically developing children as they learned to control a robotic arm. Subjects attempted to reach a target of interest while the robotic arm produced a force perpendicular to that location.

To test this hypothesis, children engaged in a second experiment in which Dr. Mostofsky’s group observed and assessed generalization, the signature of activation fields of neurons. “Generalization can tell you about how [children] learn, because you can look at the way they are able to transfer what they learn in one particular state to another,” Dr. Mostofsky said.

“The generalization patterns were strikingly different,” Dr. Mostofsky and colleagues reported. Typically developing children generalized in both proprioceptive and visual coordinates when generating models of behavior; whereas, children with autism spectrum disorders only generalized in proprioceptive coordinates, and approximately twice as strong as the typically developing children. Furthermore, the tendency to generalize in proprioceptive coordinates was highly predictive of autism-associated impairments in performance in skilled motor gestures to imitation, as well as performance of these gestures to command, and with actual tool use (often referred to as “dyspraxia”).

“[Moreover], notions of feed-forward hypotheses would suggest that these same internal models that are the basis of learning skilled movements might also be the basis for which our brain learns to understand and recognize the actions of others,” Dr. Mostofsky stated. Therefore, impaired acquisition of skill movements may contribute to the social deficits associated with autism.

Serum IL-6 Levels

Consistent with this hypothesis, generalization in intrinsic proprioceptive coordinates was highly predictive of higher (more impaired) Autism Diagnostic Observation Schedule scores for children with autism, and predictive of higher (more impaired) Social Responsiveness Scale scores for children with autism and in typically developing children, according to Dr. Mostofsky.

Dr. Mostofsky and colleagues are now examining whether these findings are specific to autism. In addition, they want to determine whether the formation of internal models of action are associated with abnormal patterns of neural connectivity. “Our preliminary diffusion tensor imaging findings do suggest that disorganization of white matter in the primary sensorimotor cortex may be associated with this increased proprioceptive bias,” Dr. Mostofsky commented.

The researchers also want to determine whether these observations can be used to modify the learning patterns in autism, either on a behavioral level, or as cortical stimulation methods used to upregulate visual-premotor connections.

—Laura Sassano

LOUISVILLE—The autistic brain builds a stronger-than-normal association between motor commands and proprioceptive feedback, which may account for why children with autism have difficulty forming the models necessary to engage not only in motor behavior, but in social and communicative behaviors, according to research presented at the 38th National Meeting of the Child Neurology Society.

Stewart H. Mostofsky, MD, Associate Professor of Neurology at Kennedy Krieger Institute and the Johns Hopkins University School of Medicine in Baltimore, and colleagues, observed 14 children with autism spectrum disorders and 13 typically developing children as they learned to control a robotic arm. Subjects attempted to reach a target of interest while the robotic arm produced a force perpendicular to that location.

To test this hypothesis, children engaged in a second experiment in which Dr. Mostofsky’s group observed and assessed generalization, the signature of activation fields of neurons. “Generalization can tell you about how [children] learn, because you can look at the way they are able to transfer what they learn in one particular state to another,” Dr. Mostofsky said.

“The generalization patterns were strikingly different,” Dr. Mostofsky and colleagues reported. Typically developing children generalized in both proprioceptive and visual coordinates when generating models of behavior; whereas, children with autism spectrum disorders only generalized in proprioceptive coordinates, and approximately twice as strong as the typically developing children. Furthermore, the tendency to generalize in proprioceptive coordinates was highly predictive of autism-associated impairments in performance in skilled motor gestures to imitation, as well as performance of these gestures to command, and with actual tool use (often referred to as “dyspraxia”).

“[Moreover], notions of feed-forward hypotheses would suggest that these same internal models that are the basis of learning skilled movements might also be the basis for which our brain learns to understand and recognize the actions of others,” Dr. Mostofsky stated. Therefore, impaired acquisition of skill movements may contribute to the social deficits associated with autism.

Serum IL-6 Levels

Consistent with this hypothesis, generalization in intrinsic proprioceptive coordinates was highly predictive of higher (more impaired) Autism Diagnostic Observation Schedule scores for children with autism, and predictive of higher (more impaired) Social Responsiveness Scale scores for children with autism and in typically developing children, according to Dr. Mostofsky.

Dr. Mostofsky and colleagues are now examining whether these findings are specific to autism. In addition, they want to determine whether the formation of internal models of action are associated with abnormal patterns of neural connectivity. “Our preliminary diffusion tensor imaging findings do suggest that disorganization of white matter in the primary sensorimotor cortex may be associated with this increased proprioceptive bias,” Dr. Mostofsky commented.

The researchers also want to determine whether these observations can be used to modify the learning patterns in autism, either on a behavioral level, or as cortical stimulation methods used to upregulate visual-premotor connections.

—Laura Sassano

In his June 2009 address to the American Medical Association, President Obama commented that “doctors feel like they are constantly looking over their shoulder for fear of lawsuits. Some doctors may feel the need to order more tests and treatments to avoid being legally vulnerable.”¹ By practicing what the President called “excessive defensive medicine,” doctors provide “more treatment rather than better care” and drive up the cost of health care ( Box ).^2-7

This column takes a look at how defensive practices can make psychiatric care more costly and less effective, by answering these questions:

• What is defensive medicine?
• How much medical practice is “defensive,” and what does it cost?
• Do psychiatrists practice defensive medicine?
• Does defensive psychiatric practice lead to suboptimal care?
• Are some defensive practices justified?
• Can you balance good defense with good care?

What is defensive medicine?

In a 1994 study, the U.S. Office of Technology Assessment (OTA) said that defensive medicine occurs “when doctors order tests, procedures, or visits, or avoid high-risk patients or procedures, primarily (but not necessarily or solely) to reduce their exposure to malpractice liability.” This definition does not require that defensive clinical practices provide no benefit to patients, only that the expected benefits are small relative to their costs.⁸

Preventing the worst outcome

Studies suggest that doctors develop and maintain practice habits—consciously or not—that aim to reduce their risk of getting sued for malpractice. For example, when patients presenting with tick bites express concern about Lyme disease, doctors overuse tests and needlessly prescribe antibiotics.⁹ Although these practices are not evidence-based, they reduce doctors’ anxiety by “preventing the worst outcome at relatively little risk and cost and avoiding a potential lawsuit at the same time.”¹⁰

The OTA estimated that up to 8% of diagnostic procedures were ordered primarily because of conscious concern about malpractice liability, based on physicians’ responses to a set of written scenarios.⁸ In a recent study, 83% of Massachusetts physicians reported practicing defensive medicine and estimated that defensive reasons accounted for why they ordered:

• 18% of lab tests
• up to 30% of procedures and consultations
• 13% of hospitalizations.¹¹

Almost all high-liability specialists (such as emergency room physicians, surgeons, and obstetrician/gynecologists) report practicing defensive medicine, often gaging in “assurance behavior”—ordering tests, doing diagnostic procedures, and referring patients to consultants.¹²

Defensive psychiatry

Compared with other specialists, psychiatrists are at lower risk for being sued, but we engage in defensive practices nonetheless. A survey of British psychiatrists found that during the previous month, 75% made clinical decisions—such as “overcautiously” admitting patients or ordering special observation—because of worries about possible legal claims, complaints, or disciplinary action.¹³

Younger psychiatrists and psychiatrists who have experienced complaints and critical incidents are more likely to practice defensive medicine. This is hardly surprising—a malpractice suit can be very stressful.¹⁴ But an amorphous dread of lawsuits affects many psychiatrists, including residents who never have been sued. The result: many needless, countertherapeutic, defensive practices.^15,16

Unintended consequences

Defensive medicine is not just expensive and wasteful. It could increase your risk of litigation if practices result in harm.¹⁷ Simon and Shuman¹⁶ give examples of how attempts to avoid litigation can compromise clinical care when treating patients at risk for suicide:

• not prescribing clozapine—a treatment known to lower the risk of suicide¹⁸ —to a chronically suicidal patient with schizophrenia because of fears of agranulocytosis (see “Clozapine for schizophrenia: Life-threatening or life-saving treatment?” Current Psychiatry, December 2009)
• not recommending electroconvulsive therapy—and possibly prolonging the period when a severely depressed patient is at high risk for suicide—to avoid a lawsuit related to memory loss
• hospitalizing a patient at chronic risk for suicide who could be managed as an outpatient with appropriate safeguards, a practice that could undermine a valuable treatment alliance.

 

Good clinical care lowers the likelihood of harm to patients, making it a sound risk management practice, though not a complete strategy. Even the best doctors can start to think defensively when confronted with awkward, troubling, or life-threatening situations that could have medicolegal implications.¹⁶ For example, when an outpatient threatens to hurt someone else, it may be tempting to just confine him in a hospital (which reduces the doctor’s anxiety) even when other less coercive and more therapeutic options might better resolve the patient’s problems and the risk of violence.

Recognizing that you’re making clinical decisions out of fear of getting sued is the first step toward curtailing needlessly defensive practice. See Table ¹⁹ for more strategies.

Table

3 strategies for avoiding needless defensive medicine

Ask yourself, “If I weren’t worried about getting sued, what would I do?” or “If I were my patient, what would I want me to do?” These questions, which help you identify the best clinical response, also may help you to implement it without taking extraneous defensive measures.
“Never worry alone.” This recommendation from the Massachusetts General Hospital and McLean Hospital training programs19 means that if you’re concerned about a case, ask a colleague for a consultation. In addition to being helpful and reassuring, an outside perspective can support nondefensive, patient-oriented decision making.
If the treatment course you think is best involves a legal matter, make sure you understand the legal issues. For example, civil commitment is often the right intervention for a mentally ill person who poses a serious risk of harm, but some patients threaten to sue doctors who propose involuntary hospitalization. Your hospital’s attorney may provide explanation and legal guidance if you do not thoroughly understand legal mechanisms or whether you are properly invoking them

Justifiable defensiveness

Of course, it’s perfectly appropriate for psychiatrists to recognize malpractice risks and take appropriate measures to avoid successful lawsuits. For example, thoughtful documentation of your data gathering, decision making, and informed consent is an appropriate protective practice. Usually, no one sees the documentation, and it contributes little to your patients’ well-being. Good documentation can be inexpensive, however, and if done creatively, can improve data recording that in turn contributes to better treatment.²⁰

In his June 2009 address to the American Medical Association, President Obama commented that “doctors feel like they are constantly looking over their shoulder for fear of lawsuits. Some doctors may feel the need to order more tests and treatments to avoid being legally vulnerable.”¹ By practicing what the President called “excessive defensive medicine,” doctors provide “more treatment rather than better care” and drive up the cost of health care ( Box ).^2-7

This column takes a look at how defensive practices can make psychiatric care more costly and less effective, by answering these questions:

• What is defensive medicine?
• How much medical practice is “defensive,” and what does it cost?
• Do psychiatrists practice defensive medicine?
• Does defensive psychiatric practice lead to suboptimal care?
• Are some defensive practices justified?
• Can you balance good defense with good care?

What is defensive medicine?

In a 1994 study, the U.S. Office of Technology Assessment (OTA) said that defensive medicine occurs “when doctors order tests, procedures, or visits, or avoid high-risk patients or procedures, primarily (but not necessarily or solely) to reduce their exposure to malpractice liability.” This definition does not require that defensive clinical practices provide no benefit to patients, only that the expected benefits are small relative to their costs.⁸

Preventing the worst outcome

Studies suggest that doctors develop and maintain practice habits—consciously or not—that aim to reduce their risk of getting sued for malpractice. For example, when patients presenting with tick bites express concern about Lyme disease, doctors overuse tests and needlessly prescribe antibiotics.⁹ Although these practices are not evidence-based, they reduce doctors’ anxiety by “preventing the worst outcome at relatively little risk and cost and avoiding a potential lawsuit at the same time.”¹⁰

The OTA estimated that up to 8% of diagnostic procedures were ordered primarily because of conscious concern about malpractice liability, based on physicians’ responses to a set of written scenarios.⁸ In a recent study, 83% of Massachusetts physicians reported practicing defensive medicine and estimated that defensive reasons accounted for why they ordered:

• 18% of lab tests
• up to 30% of procedures and consultations
• 13% of hospitalizations.¹¹

Almost all high-liability specialists (such as emergency room physicians, surgeons, and obstetrician/gynecologists) report practicing defensive medicine, often gaging in “assurance behavior”—ordering tests, doing diagnostic procedures, and referring patients to consultants.¹²

Defensive psychiatry

Compared with other specialists, psychiatrists are at lower risk for being sued, but we engage in defensive practices nonetheless. A survey of British psychiatrists found that during the previous month, 75% made clinical decisions—such as “overcautiously” admitting patients or ordering special observation—because of worries about possible legal claims, complaints, or disciplinary action.¹³

Younger psychiatrists and psychiatrists who have experienced complaints and critical incidents are more likely to practice defensive medicine. This is hardly surprising—a malpractice suit can be very stressful.¹⁴ But an amorphous dread of lawsuits affects many psychiatrists, including residents who never have been sued. The result: many needless, countertherapeutic, defensive practices.^15,16

Unintended consequences

Defensive medicine is not just expensive and wasteful. It could increase your risk of litigation if practices result in harm.¹⁷ Simon and Shuman¹⁶ give examples of how attempts to avoid litigation can compromise clinical care when treating patients at risk for suicide:

• not prescribing clozapine—a treatment known to lower the risk of suicide¹⁸ —to a chronically suicidal patient with schizophrenia because of fears of agranulocytosis (see “Clozapine for schizophrenia: Life-threatening or life-saving treatment?” Current Psychiatry, December 2009)
• not recommending electroconvulsive therapy—and possibly prolonging the period when a severely depressed patient is at high risk for suicide—to avoid a lawsuit related to memory loss
• hospitalizing a patient at chronic risk for suicide who could be managed as an outpatient with appropriate safeguards, a practice that could undermine a valuable treatment alliance.

 

Good clinical care lowers the likelihood of harm to patients, making it a sound risk management practice, though not a complete strategy. Even the best doctors can start to think defensively when confronted with awkward, troubling, or life-threatening situations that could have medicolegal implications.¹⁶ For example, when an outpatient threatens to hurt someone else, it may be tempting to just confine him in a hospital (which reduces the doctor’s anxiety) even when other less coercive and more therapeutic options might better resolve the patient’s problems and the risk of violence.

Recognizing that you’re making clinical decisions out of fear of getting sued is the first step toward curtailing needlessly defensive practice. See Table ¹⁹ for more strategies.

Table

3 strategies for avoiding needless defensive medicine

Ask yourself, “If I weren’t worried about getting sued, what would I do?” or “If I were my patient, what would I want me to do?” These questions, which help you identify the best clinical response, also may help you to implement it without taking extraneous defensive measures.
“Never worry alone.” This recommendation from the Massachusetts General Hospital and McLean Hospital training programs19 means that if you’re concerned about a case, ask a colleague for a consultation. In addition to being helpful and reassuring, an outside perspective can support nondefensive, patient-oriented decision making.
If the treatment course you think is best involves a legal matter, make sure you understand the legal issues. For example, civil commitment is often the right intervention for a mentally ill person who poses a serious risk of harm, but some patients threaten to sue doctors who propose involuntary hospitalization. Your hospital’s attorney may provide explanation and legal guidance if you do not thoroughly understand legal mechanisms or whether you are properly invoking them

Justifiable defensiveness

Of course, it’s perfectly appropriate for psychiatrists to recognize malpractice risks and take appropriate measures to avoid successful lawsuits. For example, thoughtful documentation of your data gathering, decision making, and informed consent is an appropriate protective practice. Usually, no one sees the documentation, and it contributes little to your patients’ well-being. Good documentation can be inexpensive, however, and if done creatively, can improve data recording that in turn contributes to better treatment.²⁰

In his June 2009 address to the American Medical Association, President Obama commented that “doctors feel like they are constantly looking over their shoulder for fear of lawsuits. Some doctors may feel the need to order more tests and treatments to avoid being legally vulnerable.”¹ By practicing what the President called “excessive defensive medicine,” doctors provide “more treatment rather than better care” and drive up the cost of health care ( Box ).^2-7

This column takes a look at how defensive practices can make psychiatric care more costly and less effective, by answering these questions:

• What is defensive medicine?
• How much medical practice is “defensive,” and what does it cost?
• Do psychiatrists practice defensive medicine?
• Does defensive psychiatric practice lead to suboptimal care?
• Are some defensive practices justified?
• Can you balance good defense with good care?

What is defensive medicine?

In a 1994 study, the U.S. Office of Technology Assessment (OTA) said that defensive medicine occurs “when doctors order tests, procedures, or visits, or avoid high-risk patients or procedures, primarily (but not necessarily or solely) to reduce their exposure to malpractice liability.” This definition does not require that defensive clinical practices provide no benefit to patients, only that the expected benefits are small relative to their costs.⁸

Preventing the worst outcome

Studies suggest that doctors develop and maintain practice habits—consciously or not—that aim to reduce their risk of getting sued for malpractice. For example, when patients presenting with tick bites express concern about Lyme disease, doctors overuse tests and needlessly prescribe antibiotics.⁹ Although these practices are not evidence-based, they reduce doctors’ anxiety by “preventing the worst outcome at relatively little risk and cost and avoiding a potential lawsuit at the same time.”¹⁰

The OTA estimated that up to 8% of diagnostic procedures were ordered primarily because of conscious concern about malpractice liability, based on physicians’ responses to a set of written scenarios.⁸ In a recent study, 83% of Massachusetts physicians reported practicing defensive medicine and estimated that defensive reasons accounted for why they ordered:

• 18% of lab tests
• up to 30% of procedures and consultations
• 13% of hospitalizations.¹¹

Almost all high-liability specialists (such as emergency room physicians, surgeons, and obstetrician/gynecologists) report practicing defensive medicine, often gaging in “assurance behavior”—ordering tests, doing diagnostic procedures, and referring patients to consultants.¹²

Defensive psychiatry

Compared with other specialists, psychiatrists are at lower risk for being sued, but we engage in defensive practices nonetheless. A survey of British psychiatrists found that during the previous month, 75% made clinical decisions—such as “overcautiously” admitting patients or ordering special observation—because of worries about possible legal claims, complaints, or disciplinary action.¹³

Younger psychiatrists and psychiatrists who have experienced complaints and critical incidents are more likely to practice defensive medicine. This is hardly surprising—a malpractice suit can be very stressful.¹⁴ But an amorphous dread of lawsuits affects many psychiatrists, including residents who never have been sued. The result: many needless, countertherapeutic, defensive practices.^15,16

Unintended consequences

Defensive medicine is not just expensive and wasteful. It could increase your risk of litigation if practices result in harm.¹⁷ Simon and Shuman¹⁶ give examples of how attempts to avoid litigation can compromise clinical care when treating patients at risk for suicide:

• not prescribing clozapine—a treatment known to lower the risk of suicide¹⁸ —to a chronically suicidal patient with schizophrenia because of fears of agranulocytosis (see “Clozapine for schizophrenia: Life-threatening or life-saving treatment?” Current Psychiatry, December 2009)
• not recommending electroconvulsive therapy—and possibly prolonging the period when a severely depressed patient is at high risk for suicide—to avoid a lawsuit related to memory loss
• hospitalizing a patient at chronic risk for suicide who could be managed as an outpatient with appropriate safeguards, a practice that could undermine a valuable treatment alliance.

 

Good clinical care lowers the likelihood of harm to patients, making it a sound risk management practice, though not a complete strategy. Even the best doctors can start to think defensively when confronted with awkward, troubling, or life-threatening situations that could have medicolegal implications.¹⁶ For example, when an outpatient threatens to hurt someone else, it may be tempting to just confine him in a hospital (which reduces the doctor’s anxiety) even when other less coercive and more therapeutic options might better resolve the patient’s problems and the risk of violence.

Recognizing that you’re making clinical decisions out of fear of getting sued is the first step toward curtailing needlessly defensive practice. See Table ¹⁹ for more strategies.

Table

3 strategies for avoiding needless defensive medicine

Ask yourself, “If I weren’t worried about getting sued, what would I do?” or “If I were my patient, what would I want me to do?” These questions, which help you identify the best clinical response, also may help you to implement it without taking extraneous defensive measures.
“Never worry alone.” This recommendation from the Massachusetts General Hospital and McLean Hospital training programs19 means that if you’re concerned about a case, ask a colleague for a consultation. In addition to being helpful and reassuring, an outside perspective can support nondefensive, patient-oriented decision making.
If the treatment course you think is best involves a legal matter, make sure you understand the legal issues. For example, civil commitment is often the right intervention for a mentally ill person who poses a serious risk of harm, but some patients threaten to sue doctors who propose involuntary hospitalization. Your hospital’s attorney may provide explanation and legal guidance if you do not thoroughly understand legal mechanisms or whether you are properly invoking them

Justifiable defensiveness

Of course, it’s perfectly appropriate for psychiatrists to recognize malpractice risks and take appropriate measures to avoid successful lawsuits. For example, thoughtful documentation of your data gathering, decision making, and informed consent is an appropriate protective practice. Usually, no one sees the documentation, and it contributes little to your patients’ well-being. Good documentation can be inexpensive, however, and if done creatively, can improve data recording that in turn contributes to better treatment.²⁰