Vildagliptin in combination with pioglitazone improves glycaemic control in patients with type 2 diabetes failing thiazolidinedione monotherapy: a randomized, placebo-controlled study*

Authors


  • *

    This trial (NCT00099853) is registered with ClinicalTrials.gov.

Alan J. Garber, Division of Endocrinology, Baylor College of Medicine, 6550 Fannin Street, Suite 1045, Houston, TX 77030, USA
E-mail:
agarber@bcm.tmc.edu

Abstract

Aim:  The purpose of this study was to assess the efficacy and tolerability of the dipeptidyl peptidase-4 inhibitor vildagliptin in combination with the thiazolidinedione (TZD) pioglitazone in patients with type 2 diabetes (T2DM).

Methods:  This was a 24-week, multicentre, double-blind, randomized, parallel-group study comparing the effects of vildagliptin (50 or 100 mg daily) with placebo as an add-on therapy to pioglitazone (45 mg daily) in 463 patients with T2DM inadequately controlled by prior TZD monotherapy. Between-treatment comparisons of efficacy parameters were made by analysis of covariance model.

Results:  The adjusted mean change (AMΔ) in haemoglobin A1c from baseline to endpoint was −0.8 ± 0.1% (p = 0.001 vs. placebo) and −1.0 ± 0.1% (p < 0.001 vs. placebo), respectively, in patients receiving vildagliptin 50 or 100 mg daily. Relative to baseline, vildagliptin added to pioglitazone also reduced fasting plasma glucose (FPG) (AMΔ FPG =−0.8 ± 0.2 mmol/l and −1.1 ± 0.2 mmol/l; not significant (NS) vs. placebo) and postprandial glucose (PPG) [AMΔ PPG =−1.9 ± 0.6 mmol/l and −2.6 ± 0.6 mmol/l (p = 0.008 vs. placebo)] for 50 and 100 mg daily respectively. Relative to placebo, both doses of vildagliptin significantly increased the insulin secretory rate/glucose by more than threefold. The overall incidence of adverse events (AEs) was 55.5, 50.0 and 48.7% in patients receiving vildagliptin 50 mg, 100 mg daily or placebo respectively. Serious AEs were experienced by 6.8, 1.3 and 5.7% of patients receiving vildagliptin 50 mg, 100 mg daily or placebo respectively. Mild hypoglycaemia was reported by 0, 0.6 and 1.9% of patients, respectively, receiving vildagliptin 50 mg, 100 mg daily or placebo.

Conclusion:  Vildagliptin is effective and well tolerated when added to a maximum dose of pioglitazone, without increasing the risk of hypoglycaemia.

Introduction

Vildagliptin is a potent and selective inhibitor of dipeptidyl peptidase-4 (DPP-4), the enzyme responsible for the rapid degradation of circulating glucagon-like peptide-1. Mechanistic studies have shown that vildagliptin improves islet function in patients with type 2 diabetes (T2DM) by increasing both α- and β-cell responsiveness to glucose [1,2]. In phase II studies, vildagliptin was shown to decrease haemoglobin A1c (HbA1c) when given as monotherapy [3,4] or when added to metformin [5], and preclinical work with vildagliptin and pioglitazone suggested that the effects of this DPP-4 inhibitor would be complementary to those of a thiazolidinedione (TZD) [6]. However, the efficacy and tolerability of vildagliptin combined with a TZD in patients with T2DM remain to be established.

Therefore, the present 24-week, multicentre, randomized, controlled clinical trial was conducted to ascertain the efficacy and tolerability of vildagliptin (50 or 100 mg daily) added to a maximum dose of pioglitazone (45 mg daily) in patients with T2DM inadequately controlled with TZD monotherapy. Mechanistic aspects were also explored during standard meal tests conducted in a subset of patients.

Research design and methods

Study design

This was a 24-week, double-blind, randomized, placebo-controlled, parallel-group study conducted at 123 centres in the USA (99) and Romania (24) in patients with T2DM inadequately controlled with TZD monotherapy. All potential study patients attended one screening visit (week–4) during which the inclusion/exclusion criteria were assessed. All eligible patients received pioglitazone at a dose of 45 mg daily (given qd) and were randomized 4 weeks later at baseline (visit 2, week 0) to receive vildagliptin 50 mg daily (as a qd dose), 100 mg daily (as equally divided doses) or placebo. Efficacy and tolerability were assessed during four additional visits at weeks 4, 12, 16 and 24 of treatment.

Study population

The study enrolled patients with T2DM who had been treated with TZD monotherapy for at least 3 months with a stable dose of at least 4 mg of rosiglitazone or 30 mg of pioglitazone for the past 4 weeks. Starting at visit 1 (week–4), all patients then received pioglitazone 45 mg daily. Eligibility criteria were age 18–80 years, body mass index 22–45 kg/m2, HbA1c 7.5–11% and fasting plasma glucose (FPG) <15 mmol/l.

Patients were excluded if they had a history of type 1 or secondary forms of diabetes, myocardial infarction, unstable angina or coronary artery bypass surgery within the previous 6 months. Congestive heart failure, liver diseases, such as cirrhosis or chronic active hepatitis, or use of any oral antidiabetic drug other than a TZD within the past 3 months also precluded participation. Patients with any of the following laboratory abnormalities were also excluded: alanine amino-transferase (ALT) or aspartate amino-transferase (AST) >2.5 times the upper limit of normal (ULN); direct bilirubin >1.3 times the ULN; serum creatinine levels >220 μmol/l, clinically significant abnormal thyroid stimulating hormone (TSH) or fasting triglycerides (TG) >7.9 mmol/l.

Study assessments

HbA1c, FPG, body weight and vital signs were measured at each study visit. Fasting insulin and proinsulin levels were measured at baseline, week 12 and week 24. Standard haematology and biochemistry laboratory assessments were made at each visit except week 16. Fasting lipid levels (TG, total, LDL, HDL, non-HDL and VLDL cholesterol) were measured and ECGs were performed at screening and at weeks 0, 12 and 24. Standard breakfast meal tests (500 kcal; 60% carbohydrate, 30% fat, 10% protein) were performed at week 0 and week 24 for assessment of prandial glucose control and β-cell function in patients agreeing to participate. Insulin secretory rate (ISR) was calculated by deconvolution of plasma C-peptide levels [7]. The 2-h area under the curve (AUC) for ISR and glucose were calculated using the trapezoidal method, and the ratio of ISR AUC to glucose AUC was used as a measure of β-cell function.

All adverse events (AEs) were recorded and assessed by the investigator as to severity and possible relationship to study medication. Patients were provided with glucose monitoring devices and supplies and instructed on their use. Hypoglycaemia was defined as symptoms suggestive of low blood glucose confirmed by self-monitored blood glucose measurement <3.1 mmol/l plasma glucose equivalent. Severe hypoglycaemia was defined as any episode requiring the assistance of another party.

All laboratory assessments were made by central laboratories. All assessments except HbA1c were performed by Bioanalytical Research Corporation (BARC-US, Lake Success, NY,USA, for the US centres and BARC-EU, Ghent, Belgium, for Romania centres). Assays were performed according to standardized and validated procedures according to Good Laboratory Practice. HbA1c measurements were performed by either BARC-EU for Romania or by Diabetes Diagnostics Laboratory (Columbia, MO, USA) or Covance-US (Indianapolis, IN, USA) for the USA. All samples from any single patient were measured by the same laboratory.

Data analysis

The primary efficacy variable was the change from baseline in HbA1c at study endpoint using last observation carried forward for patients who discontinued early. Secondary efficacy parameters included FPG, fasting plasma lipids and body weight. The primary efficacy analyses were performed with data from patients who (i) had a screening HbA1c value ≥7.4%, (ii) received at least one dose of study medication and (iii) had a baseline and at least one postbaseline HbA1c measurement. This population is referred to as the primary intent to treat (ITT) population and was prespecified as the main efficacy population. Changes from baseline in primary and secondary endpoints were analysed using an analysis of covariance model, with treatment and pooled centre as the classification variables and baseline value as the covariate. Analyses were carried out using two-tailed tests and a statistical significance level of 0.05. For HbA1c and FPG, multiple testing was adjusted using Hochberg’s multiple testing step-up procedure to maintain an overall two-sided significance level of 0.05 [8]. The data reported for safety and tolerability included all patients exposed to at least one dose of study drug and had at least one postbaseline safety assessment.

Ethics and good clinical practice

All participants provided written informed consent. The protocol was approved by the independent ethics committee/institutional review board at each study site, and the study was conducted in accordance with the Declaration of Helsinki, using Good Clinical Practice.

Results

Patients studied

Patient disposition from screening through study endpoint is summarized in figure 1, and baseline demographic and metabolic characteristics of the primary ITT population are reported in table 1. A total of 463 patients were randomized; 398 patients comprised the primary ITT population, and approximately 80% of patients in each treatment group completed the study. In the primary ITT population, the groups were well balanced at baseline, with HbA1c averaging 8.7% and FPG averaging 10.1 mmol/l in the combined cohort. Participants were predominantly Caucasian and obese, with a mean age of 54 years and mean disease duration of 4.7 years. Patients had been using a TZD for an average of ∼15–17 months, with 28–39% of participants receiving the maximum dose of TZD prior to enrolment. Standard breakfast meal tests were performed at week 0 and week 24 in a subgroup of 140 patients (approximately one-third of patients in each treatment group) with characteristics representative of the whole population.

Figure 1.

Flow of participants from screening through study endpoint. pio, pioglitazone; Vilda, vildagliptin.

Table 1.  Patients studied and baseline characteristics of the primary ITT population
 Vildagliptin 50 mg dailyVildagliptin 100 mg dailyPlacebo
  1. BMI, body mass index; FPG, fasting plasma glucose; HbA1c, haemoglobin A1c; TZD, thiazolidinedione.

Analysis populations
 Randomized population147158158
 Safety population146158158
 Primary ITT population124136138
Primary ITT population, mean ± s.d. or n (%)
 Age54.0 ± 8.254.0 ± 9.254.8 ± 10.6
 Sex
  Male68 (54.8)61 (44.9)70 (50.7)
  Female56 (45.2)75 (55.1)68 (49.3)
 Race
  Caucasian104 (83.9)108 (79.4)108 (78.3)
  Hispanic or Latino12 (9.7)12 (8.8)10 (7.2)
  Black6 (4.8)11 (8.1)13 (9.4)
  All other2 (1.6)5 (3.7)7 (5.1)
 BMI (kg/m2)32.6 ± 5.032.2 ± 5.832.3 ± 5.8
 HbA1c (%)8.6 ± 1.08.7 ± 1.28.7 ± 1.2
 FPG (mmol/l)10.3 ± 2.910.0 ± 3.310.1 ± 3.0
 Disease duration (years)4.7 ± 4.34.6 ± 4.84.8 ± 4.6
 Duration of TZD use (months)17.3 ± 20.814.9 ± 17.614.5 ± 16.1
 On maximum TZD before study39 (31.5)38 (27.9)54 (39.1)

Efficacy

All reported efficacy data are derived from the primary ITT population. The time-course of mean HbA1c during 24-week treatment with vildagliptin 50 mg daily, 100 mg daily or placebo added to pioglitazone is depicted in figure 2. The mean baseline HbA1c was 8.6 ± 0.1% in patients receiving vildagliptin 50 mg daily and 8.7 ± 0.1% in those receiving vildagliptin 100 mg daily or placebo. Mean HbA1c decreased progressively from baseline to week 16 in patients receiving either dose of vildagliptin added to pioglitazone and remained stable thereafter. There was also a modest decrease in mean HbA1c in the placebo group during the same period. Thus, at week 24, HbA1c averaged 7.7 ± 0.1%, 7.5 ± 0.1% and 8.1 ± 0.1% in patients who received vildagliptin 50 mg daily, 100 mg daily and placebo respectively. The adjusted mean change (AMΔ) in HbA1c from baseline to endpoint was −0.3 ± 0.1% in patients receiving placebo added to pioglitazone. The AMΔ HbA1c was −0.8 ± 0.1% in patients receiving vildagliptin 50 mg daily (p = 0.001 vs. placebo) and −1.0 ± 0.1% in patients receiving vildagliptin 100 mg daily added to pioglitazone (p < 0.001 vs. placebo).

Figure 2.

Mean (±s.e.) HbA1c during 24-week treatment with vildagliptin 50 mg daily (open triangles, n = 124 at baseline, n = 107 at week 24), 100 mg daily (closed triangles, n = 136 at baseline, n = 112 at week 24) or placebo (open circles, n = 138 at baseline, n = 111 at week 24) added to pioglitazone (45 mg daily) in patients with type 2 diabetes. HbA1c, haemoglobin A1c.

At study endpoint, a significantly greater proportion of patients receiving vildagliptin 50 mg daily (28.7%, p = 0.007) or 100 mg daily (36.4%, p < 0.001) achieved target HbA1c < 7% than placebo (14.8%). In addition, in patients with initial HbA1c levels ≤ 8.0%, target HbA1c was achieved by a substantially higher percentage of patients. In this subgroup of patients with better glycaemic control while on TZD monotherapy (n = 132, mean baseline HbA1c 7.6%), target HbA1c was achieved by 53.7% of patients receiving vildagliptin 50 mg daily (p = 0.06 vs. placebo) and by 68.2% of those receiving 100 mg daily (p = 0.001 vs. placebo) vs. 34.0% of patients receiving placebo.

Fasting insulin, proinsulin, FPG, lipids and body weight

Fasting insulin did not change significantly from baseline in any treatment group, and there were no significant between-treatment differences. Fasting proinsulin increased modestly in patients receiving placebo added to pioglitazone (AMΔ= 3.1 ± 2.0 pmol/l) and decreased in both vildagliptin treatment groups. Thus, relative to placebo, fasting proinsulin levels were significantly decreased, both in patients receiving vildagliptin 50 mg daily (−7.8 ± 2.9 pmol/l, p = 0.008) and those receiving 100 mg daily (−8.3 ± 2.9 pmol/l, p = 0.004).

Mean FPG decreased from baseline to endpoint in each treatment group. The AMΔ FPG was −0.8 ± 0.2 mmol/l and −1.1 ± 0.2 mmol/l, respectively, in patients receiving vildagliptin 50 mg and vildagliptin 100 mg daily. However, these reductions were not significantly different from the AMΔ FPG in patients receiving placebo (−0.5 ± 0.2 mmol/l, p = 0.025) after adjusting for multiple comparisons. Relative to baseline, no fasting lipid parameter changed by more than 9% in any of the three treatment arms. Relative to placebo, no consistent or dose-related changes were observed in patients adding vildagliptin to pioglitazone.

Patients receiving placebo experienced an AMΔ in body weight of +1.4 ± 0.3 kg. There was no additional weight gain with vildagliptin 50 mg daily (between-treatment difference = 0.1 ± 0.4 kg, p = 0.849). In patients receiving vildagliptin 100 mg daily added to pioglitazone, the between-treatment difference in AMΔ body weight was 1.3 ± 0.4 kg (p = 0.003 vs. placebo).

Prandial glucose and β-cell function

Several parameters describing postprandial glucose (PPG) and β-cell function were derived from standard breakfast meal tests performed at baseline and endpoint in approximately one-third of study participants, with comparable numbers of patients from each treatment arm. Prandial insulin levels did not change significantly from baseline in any treatment group, and there were no significant between-treatment differences. However, PPG was decreased and β-cell function, as assessed by insulin secretory rate relative to glucose (ISR/glucose), was significantly improved in patients receiving vildagliptin added to pioglitazone. Figure 3 depicts the AMΔ in 2-h PPG (panel A) and the ISR/glucose AUC (panel B) in patients receiving vildagliptin 50 mg daily, 100 mg daily or placebo added to pioglitazone. At baseline, the 2-h PPG averaged 14.4 ± 0.7 mmol/l (vildagliptin 50 mg daily), 14.3 ± 0.7 mmol/l (vildagliptin 100 mg daily) and 14.2 ± 0.6 mmol/l (placebo). The between-treatment difference in the AMΔ 2-h PPG was −1.2 ± 0.7 mmol/l in patients receiving vildagliptin 50 mg daily (p = 0.090 vs. placebo) and −1.9 ± 0.7 mmol/l in those receiving vildagliptin 100 mg daily added to pioglitazone (p = 0.008 vs. placebo).

Figure 3.

AMΔ (±s.e.) in 2-h PPG (panel A) and β-cell function (panel B) after 24-week treatment with vildagliptin 50 mg daily (hatched bars), 100 mg daily (closed bars) or placebo (open bars) added to pioglitazone (45 mg daily) in patients with T2DM. *p < 0.05; **p < 0.01 vs. placebo. For 2-h PPG (panel A), n = 48, 49 and 42 for vildagliptin 50 mg daily, 100 mg daily and placebo respectively. For β-cell function (panel B), n = 48, 48 and 42 for vildagliptin 50 mg daily, 100 mg daily and placebo respectively. AMΔ, adjusted mean change; AUC, area under the curve; ISR, insulin secretory rate; pio, pioglitazone; PPG, postprandial glucose; Vilda, vildagliptin.

At baseline, the ISR/glucose AUC (in pmol/min/m2/mmol/l) averaged 17.7 ± 1.2 (in patients receiving vildagliptin 50 mg daily), 19.1 ± 1.2 (with vildagliptin 100 mg daily) and 20.8 ± 1.5 (with placebo). Relative to placebo, either dose of vildagliptin significantly increased the ISR/glucose AUC by greater than threefold.

Safety and tolerability

Table 2 summarizes the overall AE experience and the most commonly reported specific AEs. During 24-week treatment, one or more AEs were reported by 55.5, 50.0 and 48.7% of patients receiving vildagliptin 50 mg daily, 100 mg daily or placebo respectively. Peripheral oedema occurred more frequently in patients receiving vildagliptin than in those receiving placebo added to pioglitazone, but the frequency was unrelated to vildagliptin dose. Dizziness, headache and nausea were somewhat more common in the vildagliptin 50 mg daily group than in the other two treatment groups, and arthralgia, increased weight and urinary tract infection were somewhat more frequent in the vildagliptin 100 mg daily group than in the other two treatment groups. Most of the reported AEs were classified as mild and not suspected to be related to study medication. Events classified as severe were experienced by 12 (8.2%), 4 (2.5%) and 8 patients (5.1%) in the vildagliptin 50 mg daily, 100 mg daily and placebo groups respectively.

Table 2.  AEs reported by more than 5% of patients in the safety population
 Vildagliptin 50 mg daily
(n = 146)
Vildagliptin 100 mg daily
(n = 158)
Placebo
(n = 158)
  1. AE, adverse event.

Any AE81 (55.5)79 (50.0)77 (48.7)
Peripheral oedema12 (8.2)11 (7.0)4 (2.5)
Arthralgia4 (2.7)8 (5.1)2 (1.3)
Urinary tract infection3 (2.1)8 (5.1)2 (1.3)
Weight increased3 (2.1)8 (5.1)3 (1.9)
Headache9 (6.2)5 (3.2)4 (2.5)
Dizziness8 (5.5)4 (2.5)5 (3.2)
Nausea8 (5.5)2 (1.3)4 (2.5)

Serious AEs were experienced by 10 (6.8%), 2 (1.3%) and 9 patients (5.7%) in the vildagliptin 50 mg daily, 100 mg daily and placebo groups respectively. AEs leading to discontinuation were experienced by 4.8, 3.2 and 2.5% of patients receiving vildagliptin 50 mg daily, vildagliptin 100 mg daily and placebo respectively. There were two cases of congestive heart failure: one, reported as an AE, in the vildagliptin 50 mg daily group and the other, reported as a serious AE, in the placebo added to pioglitazone group.

One patient (0.6%) receiving vildagliptin 100 mg daily experienced two mild hypoglycaemic events. Three patients (1.9%) receiving placebo added to pioglitazone experienced one hypoglycaemic event, and there were no hypoglycaemic episodes in patients receiving vildagliptin 50 mg daily. No severe hypoglycaemic events were reported in any group. No major changes or consistent trends over time were observed for any haematological, biochemical, urinalysis parameter or vital sign. The frequency of treatment-emergent ECG abnormalities was low and comparable in all treatment groups.

Discussion

The present work represents the first report on the effects of a DPP-4 inhibitor added to a TZD. The main findings of this study were that in patients with T2DM inadequately controlled by TZD monotherapy, addition of the DPP-4 inhibitor vildagliptin produced statistically significant and clinically meaningful reductions in HbA1c level; the combination had a good overall tolerability profile and was associated with a very low incidence of hypoglycaemia.

The efficacy of vildagliptin observed in the present study (Δ HbA1c=−0.8% with 50 mg and −1.0% in the 100 mg daily dose group) is generally consistent with that observed with vildagliptin monotherapy [9] and that seen when vildagliptin is added to metformin [5,10]. The modest decrease in HbA1c seen in patients receiving placebo is likely the result of maximizing the dose of pioglitazone in a significant number of patients at visit 1 (i.e. week 4). These patients may have experienced some therapeutic benefit from this dose escalation.

Vildagliptin added to pioglitazone was generally well tolerated; the frequency of any specific AE was generally low and most AEs were considered to be mild and unrelated to study medication. Serious AEs were reported for <7% of patients and premature discontinuations due to AE for <5% of patients, with the frequency and nature of AE similar between placebo and the two doses of vildagliptin studied.

Nonetheless, the increased incidence of peripheral oedema seen in both combination groups relative to placebo merits comment. Oedema is a well-known and dose-related side effect of glitazones [11,12], but vildagliptin treatment has not been associated with an increased incidence of oedema throughout an extensive clinical development programme [13,14] (and unpublished results, Novartis data on file, Novartis Pharmaceutical), and there is no mechanistic basis to expect a DPP-4 inhibitor to cause fluid retention and peripheral oedema as there is for TZDs [15,16]. While the incidence of oedema was higher in patients receiving vildagliptin 50 mg (8.2%) or 100 mg daily (7.0%) than in those receiving placebo added to pioglitazone 45 mg qd, it is well recognized that oedema is more common when a TZD is used in combination therapy [12]. Thus, an increased incidence of oedema has been reported when pioglitazone is combined with any other oral agent (e.g. 45 mg pioglitazone plus metformin: 7.6% [17], 45 mg pioglitazone plus sulphonylurea: 10.7% [17]), and the incidence of oedema is particularly high when a TZD is combined with insulin (e.g. 30 mg pioglitazone plus insulin: 17.6% [18]). It should be noted, however, that the aforementioned studies are not strictly comparable with the present study because the TZD was added to ongoing treatment rather than adding a new agent to ongoing pioglitazone.

It could be hypothesized that oedema is a correlate of the improvement in glycaemic control in the setting of TZD combination therapy, and this may be exacerbated when plasma insulin levels are elevated. TZDs may interact synergistically with insulin to cause arterial vasodilatation, leading to sodium reabsorption with a subsequent increase in extracellular volume, thereby resulting in oedema. However, it is important to note that while vildagliptin in addition to pioglitazone improved β-cell function, it did not increase absolute plasma insulin levels in the fasting state or in the postprandial period.

Weight gain is another known and dose-related side effect of TZDs [11], which is positively related to the reduction in HbA1c[19], and this is usually due in part to fluid retention and in part to increased fat deposition [20]. In the present study, with the 50 mg dose of vildagliptin, there was no weight gain above that experienced by patients receiving placebo and continuing pioglitazone monotherapy, and with a 100 mg daily dose of vildagliptin, weight gain was modest (+1.3 kg relative to placebo) and within the range usually reported for pioglitazone combined with other oral agents (e.g. 45 mg plus metformin: +2.5 kg from baseline [17]). As with oedema, weight gain tends to be more pronounced when TZDs are combined with an insulinotropic agent (e.g. 30 mg pioglitazone plus repaglinide: +5.5 kg [21]) or with insulin (e.g. 45 mg pioglitazone plus insulin: +4.1 kg [11]). Again, the absence of hyperinsulinaemia with vildagliptin treatment may explain the more modest degree of weight gain despite improved metabolic control.

Furthermore, it is very likely that the glucose dependence of vildagliptin’s effects on insulin secretion (i.e. the absence of absolute hyperinsulinaemia) explains the very low incidence of hypoglycaemia seen with the vildagliptin/pioglitazone combination despite the threefold increase in insulin secretion rate adjusted for glucose. This is not the case when a TZD is combined with an insulin secretagogue such as glimepiride [22].

The glucose-dependent enhancement of insulin secretion was shown previously using mathematical modelling of glucose and ISR measured during 24-h sampling in a mechanistic study performed in drug-naive patients with T2DM receiving vildagliptin monotherapy [2] and is consistent with the findings of the present study.

Therefore, the tolerability profile (incidence of peripheral oedema) and changes in body weight are similar to those usually reported for metformin and somewhat less than those seen with insulinotropic agents or insulin, when combined with high-dose pioglitazone. The efficacy, tolerability and low hypoglycaemic potential of vildagliptin added to pioglitazone appear to be explained in large measure by the glucose-dependent effects of vildagliptin on islet function.

In summary, the highly selective DPP-4 inhibitor vildagliptin improves glycaemic control and is well tolerated in patients with T2DM when added to a maximally effective therapeutic dose of the TZD pioglitazone.

Acknowledgements

The authors gratefully acknowledge the investigators and staff at the 123 participating centres, Bitak Bassiri for operational support and study conduct and the editorial assistance of, and helpful discussions with, Beth Dunning Lower. This study was funded by Novartis Pharmaceuticals Corporation. A list of investigators is provided in the Appendix.

Appendix

List of investigators

In Romania: Dr Mircea Munteanu, Dr Gabriela Negrisanu, Dr Milivoi Stamoran, Prof. Dr Aurel Babes, Prof. Dr Amorin Popa, Dr Gheorghe Ghise, Dr Adrian Albota, Dr Carmen Crisan, Prof. Dr Dan Cheta, Prof. Dr Radu, Lichiardopol, Dr Lavinia Pop, Dr Doina Catrinoiu, Dr Gabriela Creteanu, Dr Iosif Szilagyi, Dr Ioan Petrescu, Dr Rodica Avram, Dr Elena Aciu, Prof. Dr Mariana Graur, Dr Valerica Nafornita, Dr Alina Nicolau, Dr Remus Gagiu, Prof. Dr Maria Mota and Prof. Dr Constantin Ionescu-Tirgoviste.

In USA: Dr Dean Keriakes, Dr Eli Roth, Dr Marc Rendell, Dr Robert Lipetz, Dr Athena Philis-Tsimikas, Dr Clint Strong, Dr Neil Fraser, Dr Peter McCullough, Dr Janet McGill, Dr Arshag Mooradian, Dr Gregory Serfer, Dr Kevin Roberts, Dr Dwayne Aboud, Dr Waymon Drummond, Dr Salah El Hafi, Dr Curtis Horn, Dr Jerry Mitchell, Dr Julio Rosenstock, Dr Anicia Villafria, Dr Douglas Young, Dr Larry Doehring, Dr Robert Lapidus, Dr Marcus C. Houston, Dr Stephen Pohl, Dr Spencer B. Jones, Dr Nancy Bohannon, Dr Terence Isakov, Dr Charles B. Herring, Dr James Fidelholtz, Dr Rafael Montoro, Dr John Mallory, Dr Alan Forker, Dr Roger Cady, Dr Stephen Ong, Dr William Zigrang, Dr Ramin Farsad, Dr Neal Shealy, Dr Susan Greco, Dr Timothy M. Howard, Dr David Thorne, Dr James Allison III, Dr David Robertson, Dr Alan Garber, Dr Leann Olansky, Dr Sherwyn Schwartz, Dr George Brooks, Dr Berto Zamora, Dr James Quigley, Dr Joseph Barrera, Dr Kent Farnsworth, Dr Henry Simon, Dr Lawrence Anastasi, Dr Kenneth Hershon, Dr Matthew Portz, Dr Stephen Schneider, Dr Gregory Ledger, Dr Edward Zawada, Dr Timothy Smith, Dr Donald Loew, Dr Mark Borsheim, Dr Gregory Smith, Dr Richard Cherlin, Dr Gina Brar, Dr Thomas Littlejohn, Dr David Morin, Dr Robert Hippert, Dr Patrick Pan, Dr Kenneth Holt, Dr Scott Yates, Dr Alfred Gordon, Dr Andrew Hudnut, Dr Diane Smith, Dr James Capo Jr., Dr Ramachandran Ravichandram, Dr Addison Taylor, Dr Kenneth Hillner, Dr Peterman Prosser, Dr John Sibille, Dr Misty Zelk, Dr Harold Bays, Dr Bruce Samuels, Dr Richard Arakaki, Dr James Cato, Dr Ramon Ramirez, Dr Christopher Gibbs, Dr Ruben Pipek, Dr John Devlin, Dr David Mansfield, Dr Joseph Terrana, Dr Arthur Pitterman, Dr Angelique Barreto, Dr Kathryn R. Rigonan, Dr V. Jerome Mirkil, Dr Neil R Farris, Dr Harold Fields, Dr Robert McNeill, Dr Bruce Henson, Dr Loknath Shandilya and Dr Ralph Wade.

Ancillary