Whether vildagliptin or other agents of this class can modify underlying disease course by improving or preserving α- and β-cell function remains to be determined.
Whether vildagliptin or other agents of this class can modify underlying disease course by improving or preserving α- and β-cell function remains to be determined.
Vildagliptin is an oral antidiabetic drug that improves islet function by increasing α- and β-cell responsiveness to glucose.1 It is a member of a new pharmaceutical class that inhibits the action of DPP-4, an enzyme that degrades gut incretin hormones involved in glucose homeostasis (see Figure 1A).Vildagliptin and other DPP- 4 inhibitors such as the recently approved sitagliptin (Januvia™, manufactured by Merck & Co., Inc., Whitehouse Station, NJ) appear to act primarily by preventing rapid degradation of the incretin glucagonlike peptide-1 (GLP-1) (see Figure 1B), thereby enhancing the effects of the hormone on α- and β-cell function. In the setting of diabetes, these actions improve glucose-dependent insulin secretion and inhibition of inappropriate glucagon secretion.2-8 (A second incretin hormone, glucose-dependent insulinotropic peptide (GIP), also has powerful insulinotropic effects in healthy humans; however, patients with type 2 diabetes mellitus (T2DM) show a marked resistance to GIP action. Consequently, the hormone has not received much attention as a potential therapeutic target).3 High serum concentrations of GLP-1 may also be maintained through the use of DPP-4-resistant GLP-1 receptor antagonists, a drug class known as ‘incretin mimetics’. The first of these drugs, exenatide (Byetta®, manufactured by Amylin Pharmaceuticals, Inc., San Diego, CA), is now available. In sum, DPP-4 inhibitors have the potential both for assisting in achieving glycemic goals in anti-diabetic therapy and for modifying underlying disease course by correcting the islet cell dysfunctions characteristic of T2DM. Vildagliptin has been evaluated for anti-diabetic effects and for safety and efficacy in controlling hyperglycemia in an extensive pre-clinical and clinical trial program.
Pre-clinical Studies
Vildagliptin has been characterized as a potent, stable, selective DPP-4 inhibitor.9 The drug is slow-binding kinetically,10 which allows oral administration in onceor twice-daily dosing.5 Vildagliptin binds tightly with DPP-4, but not DPP-8 or DPP-9.11 Inactivation of these enzymes has been associated with toxic effects in rats.12 Pre-clinical studies indicated that administration of the drug to insulin-resistant obese Zucker rats 15 minutes before oral glucose challenge rapidly blocked serum DPP-4 activity by more than 90% throughout the 90-minute study and elevated GLP-1 levels by 60%, relative to untreated animals.9 Vildagliptin also improved glucose tolerance and metabolic control in a variety of animal models.13-17 Notably, mice that received vildagliptin showed improved long-term islet function and elevated β-cell secretory responses compared with animals given vehicle.18 The compound also improved glucose-stimulated insulin secretion from isolated islets. A second study confirmed these effects, reporting that vildagliptin increased β-cell number and mass and inhibited apoptosis in animal islet cells.19
Clinical Studies on Mechanisms of Anti-diabetic Effects
Initial studies of vildagliptin in patients with T2DM showed increased plasma levels of active GLP-1, improved glucose-dependent insulin secretion and β-cell function, improved insulin sensitivity, reduced inappropriate glucagon secretion, reduced fasting and post-prandial glucose, and decreased HbA1c. For example, in a four-week placebo-controlled study in 37 drug-naïve patients, vildagliptin 100mg qd significantly reduced fasting glucose, four-hour post-prandial glucose excursion, and 24-hour glucose levels.4 After a standardized breakfast, active GLP-1 levels were increased by two-fold, post-meal glucagon levels were reduced compared with placebo, and insulin levels in response to meals did not differ from placebo despite the reduced glycemia, with the latter finding suggesting improved glucose sensitivity in β-cells. The reduction in prandial glucagon levels at four weeks corresponded with improved glucose tolerance as measured by twohour glucose levels (r=0.61).
The effect of vildagliptin treatment on β-cell function was further examined in a four-week placebo controlled study 20 in 20 patients that used a mathematical model that represents insulin secretory rate as a function of:
- glucose levels (β-cell glucose dose response component);
- change in glucose over time (derivative component); and
- a potentiation factor that is a function of time and represents the potential contribution of nonglucose secretagogues or other factors to insulin secretion.
Patients received three standardized meals and underwent 24-hour sampling at baseline and on days one and 28. Vildagliptin 100mg bid significantly increased insulin secretory rate at a glucose concentration of 7 mmol/L (i.e. approximate fasting glucose level in subjects with mild T2DM), with no differences in the slope of the β-cell dose response, the derivative component, or the potentiation component compared with placebo being observed.These findings indicate that vildagliptin increases insulin secretory tone, increasing secretory rate at any given glucose level compared with the untreated state. The study also showed that glucagon levels were reduced over the entire 24-hour profile with vildagliptin. No change in oral glucose insulin sensitivity index was observed on day one, but a significant effect was observed after improvement in glycemic control at four weeks, a finding suggesting that improved sensitivity was due to improved glycemia.
Improvement or preservation of β-cell function over longer-term treatment was shown in a 52-week study in 57 metformin-treated patients who received either placebo or vildagliptin 50mg qd (a dose lower than those assessed in phase III treatment trials).21 Meal tests were performed at baseline and at 12, 24, and 52 weeks. Insulin secretion was assessed as the suprabasal C-peptide area under the curve (AUC) divided by glucose AUC during a standardized breakfast. HbA1c (baseline 7.7%) decreased in vildagliptin patients and increased in placebo/metformin patients, yielding a between-group difference of 1.1%, and fasting plasma glucose was reduced in the vildagliptin group. Insulin secretion increased in vildagliptin/metformin patients by 12 weeks and the increase was sustained over 52 weeks, whereas insulin secretion decreased in the placebo/metformin group (see Figure 2).22 Post-meal insulin sensitivity increased significantly with vildagliptin, as did insulin secretion related to insulin sensitivity (adaptation index), and the change in the adaptation index was significantly correlated with the change in HbA1c (r=-0.39).
Copyright © 2005 American Diabetes Association, from Diabetes Care,Vol.28;1936–1940.22 Reprinted with permission from the American Diabetes Association.
The potential contribution of glucagon reduction to glycemic control was suggested by a placebo-controlled single-dose study investigating the mechanism whereby vildagliptin reduces fasting plasma glucose.23 Sixteen patients received vildagliptin 100mg or placebo followed 30 minutes later (6pm) by a six-hour meal tolerance test using a double-tracer technique to assess hepatic and glucose metabolism. Endogenous glucose production was significantly suppressed after vildagliptin administration compared with placebo, whereas the rate of total glucose disposal was similar. Insulin secretion rate increased significantly after vildagliptin, and there was a significant reduction in mean plasma glucose. During the test, suppression of mean plasma glucagon was 93% greater with vildagliptin. For the entire overnight period (6pm to 8am), endogenous glucose production was reduced in the vildagliptin group, with the decrease being significantly correlated with the decrease in fasting plasma glucose measured at 8am.These findings suggest that glycemic control with vildagliptin includes a reduction in endogenous glucose production occurring in the context of both increased insulin secretion and reduced glucagon secretion. Potential disease-modifying effects of vildagliptin have been suggested by other mechanistic studies.A placebocontrolled study in 12 drug-naïve patients examined acute insulin response to glucose and insulin sensitivity using frequently sampled intravenous (IV) glucose tolerance tests performed after an overnight fast before and after 12 weeks of treatment with vildagliptin 50mg bid and after a two- to four-week washout period.24 No acute increases in active GLP-1 levels were observed during the glucose tolerance tests in patients receiving vildagliptin. Both acute insulin response to glucose and insulin sensitivity increased significantly in vildagliptin recipients at week 12, with no change in placebo patients. The disposition index (acute insulin response multiplied by insulin sensitivity) increased four-fold in vildagliptin recipients at 12 weeks, with a significant increase from baseline remaining after the two- to fourweek washout (see Figure 3). HbA1c was reduced by 0.5% in vildagliptin patients and was unchanged in placebo patients. The findings suggest that improvements in β-cell function and insulin resistance occur independent of acute increases in plasma GLP-1 and that these improvements persist in beyond cessation of vildagliptin administration. Another study using a two-step insulin infusion clamp test following a standardized mixed meal study in 16 patients receiving vildagliptin 50mg bid or placebo for six weeks indicated that extrapancreatic effects of DPP-4 inhibition account for a portion of the reduced insulin resistance observed with vildagliptin.25 A four-week placebo-controlled study in 31 patients showed that treatment with vildagliptin 50mg bid resulted in reduced eight-hour AUC and incremental AUC for total triglyceride and chylomicron triglycerides and reduced AUC for the chylomicron apolipoprotein B48 and cholesterol subfractions after a fat-rich mixed meal.
*p<0.05 versus baseline, **p< 0.05 versus placebo. From D’Alessio et al.24 Copyright © 2006 American Diabetes Association, from Diabetes,Vol.55, 2006;A108.24 Reprinted with permission from the American Diabetes Association.
These findings suggest that vildagliptin treatment improves post-prandial triglyceride-rich lipoprotein metabolism;26 as in other studies, vildagliptin also increased intact GLP-1, suppressed glucagon secretion, reduced fasting and post-prandial glucose, and reduced HbA1c (-0.4%) compared with placebo in these patients.
Phase III Treatment Trials
Large-scale, randomized, double-blind, 24- and 52- week treatment trials with vildagliptin 50mg or 100mg per day as monotherapy or combination therapy (as add-on treatment or in drug-naïve patients) in T2DM have shown that it is effective in reducing HbA1c, with greater decreases occurring with higher initial HbA1c levels, is associated with infrequent hypoglycemia, does not cause weight gain, and otherwise has a relatively benign safety profile. Thus far, details of these trials are available only in abstract form.
Monotherapy versus Placebo
A six-month placebo-controlled trial in drug-naïve patients established that vildagliptin qd or bid significantly reduces HbA1c with minimal hypoglycemia and no weight gain and is otherwise well tolerated.27 In this trial, 380 patients with HbA1c of 7.5–10% (mean age 54 years, disease duration 2.0 years, body mass index (BMI) 32.8kg/m2, HbA1c 8.4%) were included in a primary intent-to-treat analysis. Patients received vildagliptin 50mg qd (n=104), 50mg bid (n=90), or 100mg qd (n=92) or placebo (n=94). HbA1c was reduced with vildagliptin by 0.5% with 50mg qd (p=0.006), 0.5% with 50mg bid (p=0.006), and 0.6% with 100mg qd (p=0.001) compared with placebo. A total of 16% of patients had been diagnosed with diabetes within the prior two weeks; in these patients, HbA1c decreased by >1% in all study groups, including the placebo group; when only patients diagnosed >3 months before enrollment (a typical clinical trial eligibility criterion) were considered (60% of patients), the decreases versus placebo were 0.8% (p<0.001), 0.7% (p=0.003), and 0.9% (p<0.001) in the three vildagliptin dose groups, respectively. Vildagliptin was well tolerated. Mild hypoglycemia occurred in 1.2%, 0%, and 0.6% of the vildagliptin groups and in 0% of placebo patients. Body weight decreased by 0.3–1.8kg in the vildagliptin groups. Of the 605 patients that were included in the safety portion of the study, investigators found that patients receiving vildagliptin 50mg qd (n=155), 50 mg bid (n=145), and 100mg qd (n=154) experienced mean systolic/diastolic reductions of 1.4/1.0mmHg (from 131.5/80.6mmHg), 4.1/2.7mmHg (from 133.0/ 81.4mmHg), and 3.0/1.4mmHg (from 131.7/ 80.9mmHg) in the dosing groups, compared with a reduction of 1.5/0.9mmHg (from 131.3/81.2mmHg) with placebo.28Monotherapy versus Metformin
Results of a one-year comparison with metformin indicated that vildagliptin at 50mg bid maintains a clinically significant reduction in HbA1c for one year, albeit smaller than the reduction observed with metformin, does not produce weight gain or frequent hypotension over that period, and markedly reduces gastrointestinal (GI) side effects compared with metformin.29 In this trial, drug-naïve patients (mean age 53 years, disease duration 2.4 years, BMI 32.4 kg/m2, HbA1c 8.7%) were randomized to vildagliptin 50mg bid (n=526) or metformin 1,000mg bid (n=254). HbA1c rapidly decreased with both treatments (see Figure 4). Adjusted mean changes in HbA1c after one year were 1.0% with vildagliptin and 1.4% with metformin; noninferiority of vildagliptin was not established. Mild hypoglycemia occurred in 0.6% of vildagliptin patients and 0.4% of metformin patients. Body weight changes were +0.3kg with vildagliptin and -1.9kg with metformin.The overall incidence of adverse effects was 70.1% in the vildagliptin group versus 75.4% in the metformin group. GI side effects were significantly reduced in vildagliptin patients, occurring in 21.8% versus 43.7% of metformin patients (p<0.001), including diarrhea in 6.0% versus 26.2%, nausea in 3.3% versus 10.3%, abdominal pain in 2.3% versus 7.1%, dyspepsia in 1.2% versus 4.8%, and vomiting in 2.1% versus 4.4%.
: Mean HbA1c over 1 Year in Patients Receiving Metformin 1000mg Bid or Vildagliptin 50mg Bid
Copyright © 2006 American Diabetes Association, from Diabetes,Vol.55, 2006;A29.29 Reprinted with permission from the American Diabetes Association.
Monotherapy versus Rosiglitazone
A comparative trial versus rosiglitazone showed that vildagliptin produced a similar reduction in HbA1c, improved lipid measures, and did not produce weight gain compared with rosiglitazone.30 In a six-month trial, 697 drug-naïve patients (mean age 54 years, disease duration 2.4 years, BMI 32.4kg/m2, HbA1c 8.7%) were randomized to vildagliptin 50mg bid (n=459) or rosiglitazone 8mg qd (n=238). HbA1c was reduced by 1.1% with vildagliptin (see Figure 5), which was noninferior to the reduction with rosiglitazone. Among patients with baseline HbA1c >9.0%, reductions were 1.8% with vildagliptin (n=166) and 1.9% with rosiglitazone (n=88). Compared with rosiglitazone, vildagliptin significantly reduced triglycerides (-9%, p=0.010), total cholesterol (-14%, p<0.001), low-density lipoprotein cholesterol (-16%, p<0.001), non–highdensity lipoprotein cholesterol (-16%, p<0.001), and very low-density lipoprotein cholesterol (-8%, p=0.007), but was associated with a significantly smaller increase in high-density lipoprotein (HDL) cholesterol (-5%, p=0.003). Overall, adverse events occurred in 61.4% of vildagliptin patients and 64.0% of rosiglitazone patients; edema occurred in 2.5% versus 4.9%. Mild hypoglycemia (one event) occurred in one patient in each group. Body weight decreased by 0.3kg with vildagliptin and increased by 1.6kg with rosiglitazone (p<0.001 for comparison).
Copyright © 2006 American Diabetes Association, from Diabetes,Vol.55, 2006;A133.30 Reprinted with permission from the American Diabetes Association.
Addition to Metformin
The addition of vildagliptin to on-going metformin therapy produced clinically meaningful dose-related reductions in HbA1c and fasting plasma glucose and appeared to reduce metformin-related GI adverse effects.31 In this 24-week study, 416 patients on stable metformin therapy of ≥1,500mg/d (mean age 56 years, disease duration 6.2 years, BMI 32.8kg/m2, HbA1c8.4%, metformin dose 2,102mg/d) were randomized to vildagliptin 50mg qd or bid or placebo. HbA1c was reduced by 0.7% with vildagliptin 50mg qd (p<0.001) and 1.1% with 50mg bid (p<0.001) versus placebo (see Figure 6); fasting plasma glucose was reduced by 0.8mM (p=0.003) and 1.7mM (p<0.001) at the 50mg qd and 50mg bid doses. Overall, adverse events occurred in 63.3% and 65.0% of the vildagliptin/metformin groups and in 63.5% of the placebo/metformin group. Adding vildagliptin appeared to reduce the frequency of metforminrelated GI adverse effects; GI adverse events occurred in 9.6% of patients receiving vildagliptin 50mg qd (p=0.022 versus metformin alone), 14.8% receiving 50mg bid, and 18.2% receiving placebo/metformin. Hypoglycemia (one event) occurred in only one patient in each study group.
Copyright © 2006 American Diabetes Association, from Diabetes,Vol.55, 2006;A29.31 Reprinted with permission from the American Diabetes Association.
Metformin patients who added vildagliptin (50mg qd or bid) also experienced a beneficial effect on blood pressure, relative to placebo, after 24 weeks of treatment.28 Systolic/diastolic blood pressure was reduced by 2.0/0.8mmHg (from 131.2/80.2mmHg) and 3.5/2.2mmHg (from 134.1/80.8mmHg) in the two vildagliptin groups, respectively, compared with a reduction of 0.8/0.1mmHg (from 133.3/79.8mmHg) with placebo/metformin. Although suggestive, the influence of vildagliptin on blood pressure remains to be clarified.
Combination with/in Addition to Pioglitazone
Two 24-week studies examining the combination of vildagliptin and pioglitazone showed that vildagliptin produces clinically meaningful reductions in HbA1c when combined with reduced-dose pioglitazone in drug-naïve patients and when added to full-dose glitazone treatment.32 In study 1, 592 drug-naïve patients were randomized to: pioglitazone 30mg qd (n=157); vildagliptin 50mg qd plus pioglitazone 15mg qd (n=139); vildagliptin 100mg qd plus pioglitazone 30mg qd (n=146); or vildagliptin 100mg qd (n=150). Baseline HbA1c ranged from 8.6 to 8.8% in all treatment groups in both studies. As shown in Figure 7, the greatest reductions in HbA1c occurred with vildagliptin 100mg qd plus pioglitazone 30mg qd (1.9%) and with vildagliptin 50mg qd plus pioglitazone 15mg qd (1.7%). Among patients with baseline HbA1c >9.0%, those receiving vildagliptin 100mg qd plus pioglitazone 30mg qd (n=54) had a 2.8% reduction (from baseline of 10.0%). In the second study, 398 patients receiving full-dose pioglitazone (45mg qd) for at least four weeks were randomized to the addition of vildagliptin 50mg qd or 50mg bid or placebo. Both vildagliptin doses produced additional significant reductions in HbA1c (see Figure 7).
From Baron MA, et al. Diabetologia (2006);49(suppl 1): pp. 485–486.32 Reproduced with kind permission from Springer-Verlag GmbH.
All treatments in both trials were well tolerated; adverse event rates ranged from 46% in patients receiving vildagliptin 50mg qd plus pioglitazone 15mg qd to 56% in patients receiving vildagliptin 50mg qd plus pioglitazone 45mg qd.
Addition to Insulin
The addition of vildagliptin to insulin treatment produced a significant reduction in HbA1c that was particularly marked in older patients and appeared to reduce the frequency and severity of hypoglycemia.33 In a 24-week trial, 256 patients with inadequate glycemic control (HbA1c 7.5–11%) on insulin treatment of >30U/d were randomized to receive vildagliptin 50mg bid (n=125) or placebo (n=131).At baseline, patients had a mean age of 59 years, with 32% aged >65 years, mean disease duration of 14.6 years, BMI of 33.0 kg/m2,HbA1c of 8.5%, fasting plasma glucose of 9.3 mM, mean duration of insulin treatment of 76 months, and a mean daily insulin dose of 82U; patients had multiple comorbidities, and 45% were taking more than five concomitant medications. Changes in HbA1c were -0.5% in patients receiving vildagliptin and -0.2% in those receiving placebo (p=0.022); among patients ≥65 years of age, vildagliptin reduced HbA1c by 0.7% compared with no change for placebo (p<0.001). The daily insulin dose increased by 2.6U/d from a baseline of 79.6U/d in vildagliptin patients and by 5.0U/d from a baseline of 81.9U/d in placebo patients (p=NS). Hypoglycemia was less common and less severe in patients receiving vildagliptin: in the vildagliptin/insulin group, 33 patients (26%) had a total of 133 events including no severe events; in the placebo/insulin group, 45 patients (34%) had a total of 185 events including six severe events.The authors suggested that the reduced frequency and severity of hypoglycemia with vildagliptin might be related to improved β-cell glucose sensing in these insulin-treated patients.Conclusion
Vildagliptin is a member of new class of antidiabetogenic compounds designed to address the primary incretin hormones involved in glucose homeostasis. In contrast with incretin mimetics, which prolong GLP-1 circulation time, DPP-4 inhibitors block the action of DPP-4 itself. In so doing, vildagliptin and sister compounds enhance glycemic control by improving islet cell responsiveness to glucose and restoring the balance between insulin and glucagon.
Clinical trials show that vildagliptin produces substantial reductions in HbA1c when used alone or in combination in drug-naïve patients and when added to on-going antidiabetic Clinical trials show that vildagliptin produces substantial reductions in HbA1c when used alone or in combination in drug-naïve patients and when added to on-going antidiabetic therapy and that it maintains efficacy in glycemic control for at least one year. It is also suitable for oncedaily dosing. Hypoglycemia and weight gain are common drawbacks of current anti-diabetic treatments; the infrequency of hypoglycemia and the weight-neutral effect of vildagliptin are thus advantageous characteristics, both in single-agent treatment and when combined treatment is considered. Other advantageous safety/tolerability features include the significantly reduced frequency of GI adverse effects compared with metformin and the possibility that coadministration reduces these effects in metformin-treated patients, and the apparent infrequency of edema. The long-term clinical benefits of the potential effects of vildagliptin or other DPP-4 inhibitors in modifying disease course by sustaining or improving islet cell function remain to be determined, as do the consequences of such other reported effects as improved post-prandial TG-rich lipoprotein metabolism and extrapancreatic mechanisms of reducing insulin resistance.The first DPP-4 inhibitor, sitagliptin, was approved in October 2006. It remains to be determined whether differences with regard to chemical characteristics and metabolism among these agents will translate into differences in efficacy, safety, and tolerability.
