Efficacy and Safety of Vildagliptin in New-Onset Diabetes After Kidney Transplantation—A Randomized, Double-Blind, Placebo-Controlled Trial

Authors


Abstract

New-onset diabetes after transplantation (NODAT) is a serious complication after kidney transplantation, but therapeutic strategies remain underexplored. Dipeptidyl peptidase-4 (DPP-4) inhibitors selectively foster insulin secretion without inducing hypoglycemia, which might be advantageous in kidney transplant recipients (KTRs) with NODAT. We conducted a randomized, double-blind, placebo-controlled, phase II trial to assess safety and efficacy of the DPP-4 inhibitor vildagliptin. Intraindividual differences in oral glucose tolerance test (OGTT)-derived 2-h plasma glucose (2HPG) from baseline to 3 months after treatment served as primary endpoint. Among secondary outcomes, we evaluated HbA1c, metabolic and safety parameters, as well as OGTTs at 1 month after drug discontinuation. Of 509 stable KTRs who were screened in our outpatient clinic, 63 (12.4%) had 2HPG ≥200 mg/dL, 33 of them were randomized and 32 completed the study. In the vildagliptin group 2HPG and HbA1c were profoundly reduced in comparison to placebo (vildagliptin: 2HPG = 182.7 mg/dL, HbA1c = 6.1%; placebo: 2HPG = 231.2 mg/dL, HbA1c = 6.5%; both p ≤ 0.05), and statistical significance was achieved for the primary endpoint (vildagliptin: 2HPG-difference −73.7 ± 51.3 mg/dL; placebo: −5.7 ± 41.4 mg/dL; p < 0.01). Adverse events were generally mild and occurred at similar rates in both groups. In conclusion, DPP-4 inhibition in KTRs with overt NODAT was safe and efficient, providing a novel treatment alternative for this specific form of diabetes.

Abbreviations
2HPG

2-h plasma glucose

AUC

area under the curve

BMI

body mass index

CKD

chronic kidney disease

DM

diabetes mellitus

DPP-4

dipeptidyl peptidase-4

FPG

fasting plasma glucose

fsOGTT

frequent sampling OGTT

GIP

glucose-dependent insulinotropic peptide

GLP-1

glucagon-like peptide-1

GOT

glutamic oxaloacetic transaminase

IFG

impaired fasting glucose

IGT

impaired glucose tolerance

ISIcomp

composite insulin sensitivity index

KTRs

kidney transplant recipients

NODAT

new-onset of diabetes after transplantation

OGIS

oral glucose insulin sensitivity index

OGTT

oral glucose tolerance test

QUICKI

quantitative insulin sensitivity check index

TAHG

transplant associated hyperglycemia

VINODAT

Vildagliptin In New Onset Diabetes After kidney Transplantation

Introduction

Transplant associated hyperglycemia (TAHG) comprises new-onset diabetes after transplantation (NODAT), impaired fasting glucose (IFG) and/or impaired glucose tolerance (IGT), all of which are associated with increased morbidity and mortality in kidney transplant recipients (KTRs) [1-3]. NODAT confers a high risk for premature graft failure and also increased cardiovascular mortality [4, 5]. Current therapeutic strategies against NODAT are largely based on treatment guidelines for type 2 diabetes mellitus (DM) [6]. While type 2 DM is often considered a problem of increased insulin resistance, recent data demonstrate that impaired insulin secretion rather than impaired insulin sensitivity might be the principal pathophysiological defect in KTRs [7, 8]. Hence, antidiabetic drugs that protect or even improve pancreatic beta cell function may be especially advantageous in NODAT treatment [9]. While beta cell protection through early postoperative insulin administration has been shown effective against posttransplant hyperglycemia, and is thereby promising for NODAT prevention [5, 10], solid evidence for antihyperglycemic treatment of overt NODAT in stable KTRs is still scarce. A recent randomized controlled trial has evaluated the impact of the oral antidiabetic agents vildagliptin and pioglitazone compared to placebo, but importantly, in KTRs with prediabetes. Both agents were found to be safe and potentially effective in their application [11].

While metformin still remains the initial choice in the treatment of type 2 DM due to its potential pleiotropic effects and excellent safety profile, its widespread use in patients with impaired kidney function is prohibited at least in advanced chronic kidney disease (CKD) stages due to an increased risk of lactic acidosis [5, 12]. By increasing insulin sensitivity, thiazolidinediones may offer an interesting treatment alternative, although their use is problematic due to side effects such as weight gain, higher risk for congestive heart failure and osteoporosis, as well as bladder cancer [13-17]. Sulfonylureas and meglitinides, which unspecifically increase insulin secretion, do not prevent or delay the loss of islet cells in diabetics, and lead to an increased hypoglycemia rate, especially in patients with CKD [18-20]. These compounds may have an additionally detrimental impact on insulin secretion and thereby accelerate beta cell loss, especially in conditions of beta cell stress such as in NODAT [5, 8, 18].

Dipeptidyl peptidase-4 (DPP-4) inhibitors belong to a new class of antidiabetic drugs that stabilize the incretin hormones glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic peptide, resulting in improved metabolic control and reduction of postprandial hyperglycemia [21]. Importantly, incretin-based therapies may exert a beta cell protective effect [22]. Furthermore, gliptins ameliorate oxidative stress and beta cell destruction in rodents with and without diabetes [23-26]. Interestingly, studies with DPP-4 inhibitors have indicated beneficial effects on blood vessels and the heart via both GLP1-dependent and GLP1-independent effects, and GLP-1 receptor activation was recently linked to control of blood pressure [27-29]. Furthermore, GLP-1 activation has been demonstrated to protect islet cells against corticosteroid-induced apoptosis [30, 31].

Clinically, gliptins have been demonstrated to be efficient antidiabetic drugs, which reduce dose-dependently the HbA1c by 0.5–1.0% in type 2 DM and have an excellent safety profile with a hypoglycemia incidence similar to placebo level [32-35]. Of note, gliptins have also been shown to be both safe and effective in CKD, as well as in patients on maintenance hemodialysis that comprise a uniquely sensitive population with regard to hypoglycemia [36-39]. These unique characteristics along with the particular pathophysiology of NODAT [7] indicate a potential suitability of DPP-4 inhibition in the treatment of overt NODAT patients, especially those with CKD.

In the present study, we examined the effects of vildagliptin in stable KTRs with newly diagnosed NODAT compared to life style intervention alone.

Methods and Study Design

Study design and subjects

All stable patients with at least 6 months follow-up after kidney transplantation at the outpatient clinic of the Clinical Department of Nephrology and Dialysis from the Vienna General Hospital/Medical University of Vienna, Austria, were routinely assigned to a 75 g-2 h-oral glucose tolerance test (OGTT), and their data were analyzed in the context of the TAHG study, an open, noninterventional, observational cohort study (Ethics Committee approval 566/2009), described elsewhere [7]. IGT was defined as 2-h plasma glucose (2HPG) of 140–199 mg/dL during the OGTT, and IFG as fasting glucose of 100–125 mg/dL [40]. All patients who had newly diagnosed NODAT defined by 2HPG ≥200 mg/dL between March 2009 and July 2012 were invited to participate in the Vildagliptin In New Onset Diabetes After kidney Transplantation (VINODAT) study. This study was designed as a prospective 16-week, single-center, double-blind, randomized, placebo-controlled, phase II trial to assess the safety and efficacy of vildagliptin in NODAT patients [41].

Recruitment procedures are depicted in Figure 1. Inclusion criteria were transplant vintage ≥6 months, stable graft function (no clinically meaningful change of the renal function within the last 3 months) and informed consent of the patient. Exclusion criteria consisted of prior history of type 1 or type 2 DM, pregnancy, severe renal impairment with an estimated GFR ≤30 mL/min/1.73 m2 (MDRD formula), and severe liver impairment with glutamic oxaloacetic transaminase (GOT) and/or GPT levels increased at least three times above the upper reference values [41].

Figure 1.

Screening, enrollment, randomization and follow-up of study participants. The study population comprised 16 patients in each group. 2HPG, 2-h plasma glucose; OGTT, oral glucose tolerance test.

Ethical declaration

The VINODAT study was approved by the local ethics committee (EK#645/2009) and by the Austrian regulatory authority (Federal Office for Safety in Health Care, Austrian Agency for Health and Food Safety) and was registered at the European Clinical Trials Database (EUDRACT number: 2009-14405-14). Furthermore, the study was registered in a publicly available clinical trial database (NCT 00980356, http://clinicaltrial.gov). Participants were followed until study completion or until their withdrawal from the study [41].

Study design and treatments

All patients were instructed by the study investigators regarding diet and level of physical activity, according to the American Diabetes Association guidelines [42]. Study participants were randomized in a 1:1 ratio to receive either 50 mg of vildagliptin (Galvus®; Novartis Europharm Ltd, West Sussex, UK) or placebo tablets once daily. Study medication was routinely taken 30 min before breakfast. Randomization and blinding were performed by the local pharmacy department [41].

Primary and secondary endpoints

OGTTs were performed at baseline, months 3 and 4. Additional metabolic parameters were determined at each visit and included HbA1c, fasting plasma glucose (FPG), lipids, liver enzymes and serum creatinine. The primary outcome of the study was the difference in the intraindividual change in OGTT-derived 2HPG between the two study arms from baseline to 3 months. Secondary outcomes included differences between the two study groups in the intraindividual change in OGTT-derived 2HPG from baseline to 4 months, FPG, HbA1c and fasting insulin within the groups before and after treatment, rate of side-effects, change in GFR (MDRD formula), albuminuria/proteinuria, change in liver function parameters from baseline, and immunosuppressant serum levels. All assessments at 4 months were 1 month after drug discontinuation and served to analyze continuing effects of vildagliptin on glucose homeostasis.

Depending on this study's resources (finances and manpower), some nonselected patients underwent a frequent sampling OGTT (fsOGTT) with determination of glucose, insulin and C-peptide at minutes 0, 30, 60, 90 and 120, rather than determination of 2HPG alone. Of all study patients, an fsOGTT was performed in 26 patients at baseline (14 vildagliptin and 12 placebo), 31 patients at month 3 (15 vildagliptin and 16 placebo) and 28 patients at month 4 (15 vildagliptin and 13 placebo group). Insulin sensitivity in the fasting state was calculated by the quantitative insulin sensitivity check index (QUICKI), and in dynamic conditions (postprandial), by the fsOGTT-derived oral glucose insulin sensitivity index (OGIS), a figure of glucose clearance widely validated versus the glucose clamp, as well as by the Matsuda's composite insulin sensitivity index (ISIcomp). For insulin secretion, indices of beta cell function (insulinogenic index and disposition index) and hepatic insulin extraction were assessed from fsOGTTs. All the above metabolic parameters have been fully described in Pacini and Mari [43]. Areas under the curve (AUCs) for glucose, insulin and C-peptide were calculated from fsOGTTs using the trapezoidal rule.

Statistical analysis

Two different analysis sets were defined for safety and efficacy, respectively. The efficacy of vildagliptin was assessed in all subjects per protocol, that is, in all patients who received the study drug (at least one dose) and did not violate the protocol in a way that might affect the evaluation of the effect of the study drug on the primary objective. The per-protocol set was employed in the analysis of efficacy variables. The safety analysis set included subjects who were randomized and received at least one dose of the study drug (modified intention-to-treat), and was employed in the analysis of all tolerability and safety variables. A sensitivity analysis was performed for efficacy with the intention-to-treat population [41].

For the primary endpoint analysis, intraindividual differences between treatment and control group in the 2HPG value obtained during an OGTT from baseline to 3 months were assessed in the per protocol population. Using two-sided testing and a standard deviation of 20% in relative changes of 2HPG OGTT glucose values, α = 0.05 and β = 0.2, a sample size of 16 patients per group was chosen to detect a difference in 2HPG of 20 mg/dL when comparing baseline levels to levels on months 3. SPSS 19 (SPSS, Inc., Chicago, IL) was used for all statistical analyses.

Results

Prevalence of impaired glucose metabolism and patient recruitment

A total of 509 KTRs without previously described diabetes and stable kidney function ≥6 months after transplantation were routinely assigned to an OGTT between March 2009 and July 2012. Of all examined KTRs, 213 (41.8%) displayed a glucose metabolism disorder, consistent with our previous results evaluating the first 307 OGTTs, performed between March 2009 and March 2010 [7]. Specifically, 64 KTRs (12.5%) patients had IFG, 35 KTRs (6.9%) had IGT, 51 KTRs (10.0%) had IFG plus IGT and 63 KTRs (12.4%) had a diabetic OGTT.

Thirty-three patients underwent randomization; 30 patients were excluded despite a diabetic OGTT (12 patients due to medical reasons including elevated liver parameters and/or severe impaired kidney function, 9 patients because they declined to participate and 9 patients because they were already assigned to other studies). One patient withdrew informed consent before taking the first dose of medication. Thirty-two patients completed the primary endpoint analysis, but two patients from the control group declined their OGTT at month 4, and one patient was lost to follow-up (Figure 1).

Baseline clinical characteristics

Demographic and baseline clinical characteristics of the two groups are depicted in Table 1. There were no significant differences in age, sex, body mass index (BMI), serum creatinine, GFR, proteinuria, time after transplantation, lipid metabolism, liver function, hemoglobin levels and corticosteroid dose among the two groups at baseline. Furthermore, no differences in glucose metabolism (FPG, 2HPG, HbA1c or fasting insulin) could be detected. However, significant baseline differences were observed in the use of mycophenolic acid (significantly higher in the placebo group) and mycophenolate mofetil (significantly higher in the vildagliptin group).

Table 1. Baseline characteristics between the vildagliptin and placebo group
 Vildagliptin (N = 16)Placebo (N = 16)p-Value
  1. ATG, anti-thymocyte globulin; BMI, body mass index; CMV, cytomegalovirus; CsA, cyclosporin A; eGFR, estimated glomerular filtration rate (MDRD formula); Gamma-GT, Gamma-glutamyltransferase; GOT, glutamic oxaloacetic transaminase; GPT, glutamic pyruvic transaminase; HDL, high-density lipoprotein; KTx, kidney transplantation; LDL, low-density lipoprotein; RAAS, renin-angiotensin-aldosterone system.
  2. Data are presented as means ± SD or frequencies and percentages.
Men10 (62.5%)14 (87.5%)0.102
Age (year)64.25 ± 8.763.0 ± 8.40.683
Time since KTx (month)69.9 ± 63.951.4 ± 47.20.383
First KTx16 (100%)16 (100%)1.000
Deceased donors16 (100%)16 (100%)1.000
Family history of diabetes5 (31.3%)6 (37.5%)0.710
BMI (kg/m2)28.4 ± 4.527.9 ± 6.00.779
HbA1c (%)6.7 ± 0.736.7 ± 0.820.946
Fasting plasma glucose (mg/dL)129.0 ± 28.1126.8 ± 20.20.763
2-h plasma glucose (mg/dL)256.4 ± 49.8236.9 ± 32.20.386
Fasting plasma insulin (µU/ml)8.1 ± 8.4 (N = 14)7.2 ± 3.9 (N = 12)0.753
Fasting plasma C-peptide (ng/ml)4.5 ± 3.9 (N = 14)3.7 ± 1.4 (N = 12)0.503
Serum creatinine (mg/dL)1.25 ± 0.321.46 ± 0.380.109
eGFR (mL/min/1.73 m2)58.3 ± 16.353.6 ± 14.40.396
Uric acid (mg/dL)7.2 ± 1.59.2 ± 5.90.210
Urine albumin-to-creatinine ratio (mg/g)22.1 ± 32.540.6 ± 90.10.446
Hemoglobin (g/dL)13.4 ± 1.213.6 ± 2.00.653
Lipid levels (mg/dL)
Total cholesterol217.1 ± 49.1195.6 ± 32.70.155
LDL cholesterol126.9 ± 45.3109.6 ± 28.10.346
HDL cholesterol55.9 ± 23.157.0 ± 23.10.611
Triglycerides179.9 ± 78.4172.44 ± 85.90.800
GOT (U/L)24.9 ± 9.528.7 ± 19.20.489
GPT (U/L)26.0 ± 13.532.2 ± 28.80.282
Gamma-GT (U/L)53.3 ± 58.7105.7 ± 132.10.158
Serum protein (g/L)71.5 ± 4.169.6 ± 3.90.181
Blood pressure (mmHg)
Systolic131.9 ± 6.0127.9 ± 10.20.188
Diastolic72.6 ± 7.677.3 ± 9.70.140
Induction therapy (basiliximab, ATG)6 (37.5%)5 (31.3%)0.884
Immunosuppression
Tacrolimus10 (62.5%)14 (87.5%)0.102
Trough levels (ng/mL)7.74 ± 2.26.2 ± 2.20.197
CsA6 (37.5%)2 (12.5%)0.102
Trough levels (ng/mL)81.7 ± 24.673.0 ± 6.00.083
Aprednislon dose (mg/day)3.6 ± 1.84.1 ± 2.60.521
Antimetabolites
Mycophenolic acid8 (50.0%)14 (87.5%)0.022
Mycophenolate mofetil7 (43.8%)2 (12.5%)0.049
Azathioprine1 (6,2%)0 (0.0%)0.310
Lipid-lowering agents5 (31.3%)9 (56.3%)0.154
Beta blockers10 (62.5%)13 (81.3%)0.238
Calcium-channel blockers11 (68.8)10 (62.5%)0.710
RAAS inhibitors8 (50%)8 (50%)1.000
Thiazides2 (12.5%)2 (12.5%)1.000
Other diuretics2 (12.5%)6 (37.5%)0.102
Viral infections
Hepatitis C0 (0%)4 (25%)0.690
CMV11 (68.8%)11 (68.8%)1.000

Metabolic outcomes

Figure 2 shows the changes in mean metabolic parameters. The primary endpoint (difference in intraindividual change in 2HPG between the vildagliptin and placebo group from baseline to 3 months) was statistically significant (vildagliptin vs. placebo: −73.7 ± 51.3 mg/dL vs. −5.7 ± 41.4 mg/dL; p < 0.01; Figure 2B). Furthermore, the changes in HbA1c also showed a significant difference (vildagliptin vs. placebo: −0.6 ± 0.5% vs. −0.1 ± 0.5%; p = 0.016; Figure 2C), while changes in fasting glucose did not show a significant difference (vildagliptin vs. placebo: −16.4 ± 16.6 mg/dL vs. −3.3 ± 20.9 mg/dL; p = 0.081; Figure 2A). At month 4, no significant differences between the two groups could be observed for changes in fasting glucose (vildagliptin vs. placebo: −7.3 ± 21.3 mg/dL vs. 8.3 ± 29.3 mg/dL; p = 0.161), changes in 2HPG (vildagliptin vs. placebo: −23.6 ± 61.3 mg/dL vs. −40.7 ± 105.4 mg/dL; p = 0.609), and changes in HbA1c (vildagliptin vs. placebo: −0.5 ± 0.4% vs. 0.0 ± 0.8%; p = 0.081). No significant changes in hemoglobin levels were observed.

Figure 2.

Intraindividual differences in metabolic parameters from baseline to 3 months between the vildagliptin and placebo groups. Mean differences ± SD between baseline and 3 months are shown for patients in the vildagliptin and placebo group for (A) fasting plasma glucose (FPG; p = 0.08); (B) 2-h plasma glucose (2HPG; p < 0.01); (C) HbA1c (p < 0.02). The asterisk indicates findings with p < 0.05.

Overall group comparisons of the metabolic parameters at month 3 showed significant differences between the vildagliptin and placebo group for 2HPG (p = 0.034; Figure 3B), and HbA1c (p = 0.039; Figure 3C), but not for FPG (p = 0.249; Figure 3A). HbA1c was still significantly different between the vildagliptin and placebo group at month 4 (p = 0.025; Figure 3C). Overall group differences (not the individual changes) from baseline to month 3, respectively month 4, are depicted in Table S1.

Figure 3.

Metabolic parameters between the vildagliptin and placebo groups, at baseline, month 3 and month 4. (A) Fasting plasma glucose (FPG; p = 0.269); (B) 2-h plasma glucose (2HPG; p = 0.024); (C) HbA1c (p = 0.025). The asterisk indicates findings with p < 0.05.

To gain further mechanistic insights, a subset of patients underwent fsOGTT at baseline and at months 3 and 4. Table S2 shows the calculated metabolic parameters. AUC glucose was significantly lower in the vildagliptin group from baseline to month 3, but returned to baseline values after 1 month from discontinuation. No differences were observed in insulin sensitivity, both at fasting (QUICKI) and in dynamic conditions (ISIcomp, OGIS). On the other hand beta cell function improved significantly after 3 months and remained slightly higher, though not significantly, at month 4.

Lifestyle modification

All patients received instructions for lifestyle modification including structured advice for physical activity by a physician and dietary advice by a professional dietician [41]. In the placebo group, however, no improvement of glycemic metabolism or weight was observed after months 3 or 4.

Subject compliance and safety analysis

Adherence to study medication and adverse events were assessed by at least monthly interviews during and at the end of the study period. Adverse events were reversible and generally mild in nature, and are listed in Tables 2 and 3. Throughout the study period, no severe adverse events were observed, and there were no drug interactions with the immunosuppressants. Kidney function as assessed by estimated GFR (MDRD formula) was not influenced by the treatment in either of the study arms. Liver function parameters did not show any differences between the groups at months 3 or 4 (Table 2).

Table 2. Safety analysis for adverse events, kidney function and metabolic parameters
 VildagliptinPlacebo
Baseline—3 months (N = 16)p-ValueBaseline—4 months (N = 16)p-ValueBaseline—3 months (N = 16)p-ValueBaseline—4 months (N = 16)p-Value
  1. BMI, body mass index; BP, blood pressure; CsA, cyclosporin A; eGFR, estimated glomerular filtration rate (MDRD formula); Gamma-GT, Gamma-glutamyltransferase; GOT, glutamic oxaloacetic transaminase; GPT, glutamic pyruvic transaminase; HDL, high-density lipoprotein; LDL, low-density lipoprotein; tac, tacrolimus.
Adverse events3 5 3 6 
Change in eGFR (mL/min/1.73 m2)1.9 ± 10.30.4794.7 ± 10.40.1012.1 ± 6.10.2171.0 ± 9.10.748
Change in BMI (kg/m2)−0.2 ± 1.10.472−0.2 ± 1.00.149−0.1 ± 0.90.6810.5 ± 1.50.166
Change in systolic BP (mmHg)−0.4 ± 6.60.829−0.3 ± 6.60.8852.5 ± 11.30.4050.8 ± 11.20.540
Change in diastolic BP (mmHg)1.1 ± 6.40.5332.1 ± 4.90.1221.9 ± 10.10.485−2.0 ± 11.80.682
Change in lipid levels (mg/dL)
Total cholesterol−2.1 ± 29.00.780−3.9 ± 24.60.545−3.1 ± 18.50.531−9.4 ± 22.70.561
LDL cholesterol0.3 ± 11.70.377−5.6 ± 19.40.283−4.4 ± 13.70.234−3.1 ± 17.30.518
HDL cholesterol−5.7 ± 24.20.919−2.4 ± 5.30.095−5.7 ± 22.60.342−9.4 ± 22.70.142
Triglycerides4.8 ± 63.10.77418.3 ± 63.50.2837.6 ± 48.30.55317.2 ± 71.40.290
Change in hepatic parameters
GOT (U/L)1.4 ± 11.90.6470.1 ± 7.00.946−1.3 ± 12.00.678−1.8 ± 11.90.581
GPT (U/L)−1.1 ± 13.20.759−4.2 ± 10.10.1272.1 ± 19.00.671−2.5 ± 19.00.635
Gamma-GT (U/L)5.3 ± 41.80.63431.4 ± 71.50.099−6.9 ± 39.60.508−6.7 ± 54.30.650
Change in Tac levels (ng/mL)−1.2 ± 2.200.1371.3 ± 0.990.083−0.5 ± 3.990.0990.7 ± 3.020.873
Change in CsA levels (ng/mL)−3.7 ± 7.740.337−3.2 ± 8.930.464−30 ± 6.000.635−7.5 ± 12.50.174
Table 3. Adverse events
 SOCLLTVildagliptin (N = 16)Placebo (N = 16)
  1. Adverse events (AE) reported in system organ class (SOC) and low level terms (LLT) according to MedDRA14.0.
Any AE (%)  5 (31.3%)6 (37.5%)
AE leading to drug discontinuation  00
AE itemizedCardiological disordersAngina pectoris10
 InvestigationsElevated liver enzymes21
  Elevated pancreas enzymes01
 Respiratory disordersElevated triglycerides01
 Ophthalmological disordersCough01
 Renal and urinary disordersConjunctivitis10
 Hematological disordersUrinary tract infections11
  Leukopenia01

Discussion

To the best of our knowledge, this is the first randomized controlled trial evaluating a pharmacological intervention against overt NODAT in stable KTRs [5, 9]. To include the required number of study patients, we performed as many as 509 random OGTTs, and received a NODAT result in 12%. The DPP-4 inhibitor vildagliptin improved glucose metabolism, namely fasting glucose levels, 2HPG and HbA1c, at 3 months after treatment, compared to baseline. Furthermore, we also demonstrated significant improvement of both 2HPG as well as HbA1c in the vildagliptin group compared to placebo. HbA1c was still significantly lower in the vildagliptin group even 1 month after study drug withdrawal, indicating a robust improvement of glucose metabolism. Evaluation of OGTT-derived indices moreover showed improved glucose profile, due to the enhanced beta cell function with vildagliptin, but no change in insulin sensitivity in the short 3-month period. Previous studies have shown reduction of insulin resistance with vildagliptin, but after 52 weeks of treatment [44]. After 1 month of discontinuing the treatment, the great majority of parameters went back to baseline values, except the insulinogenic indices which remained slightly, though not significantly, elevated.

Compared with development of type 2 DM, NODAT development may depend on several different risk factors, as well as another metabolic pattern regarding the relationship between insulin sensitivity, insulin secretion and glycemia as its product [7]. NODAT should therefore be considered as an independent form of diabetes, which justifies stressing the novelty of our results, although DPP-4 inhibitors have previously been well characterized for the type 2 DM population. The insulin secretion problem in NODAT patients may also explain why vildagliptin had a more pronounced effect on 2HPG rather than FPG. Based on the difficulty to diagnose NODAT by FPG [45], we decided to perform OGTTs to diagnose NODAT, and furthermore, as the primary endpoint.

Because NODAT is a common and severe complication after kidney transplantation and associated with increased morbidity, mortality and graft loss [1-5], novel strategies both to prevent and treat NODAT are highly warranted. Several antidiabetic drugs used in type 2 DM, however, seem unsuitable for transplanted patients [5, 9]. For example, regarding metformin use, a prospective study in KTRs is required to address whether the risk for lactic acidosis is outweighed by beneficial effects on glucose metabolism [5, 12]. Thiazolidinediones are currently under public debate concerning increased bladder cancer and fracture risk. Nevertheless, we have previously demonstrated that pioglitazone may be safe and effective in renal patients with IGT [11].

DPP-4 inhibitors can be used even for patients on hemodialysis, and have been shown to preserve pancreatic beta cell function in animal models of diabetes [7, 38, 39]. Furthermore, DDP-4 inhibitors may preserve beta cell mass and function [46, 47], which might be among the primary goals of NODAT treatment [7, 8]. In the present study, we used fsOGTTs to demonstrate that vildagliptin improved beta cell function in stable KTRs. These findings are in accordance with a case series published by Lane et al [48], in which 15 KTRs with NODAT showed a significant improvement of glycemic control after treatment with the DDP-4 inhibitor sitagliptin for 3 months. Larger studies are now needed to assess whether sensitivity may also improve, as has been shown for patients with type 2 diabetes [44].

In contrast to other antidiabetic drugs, DPP-4 inhibitors display a low risk for hypoglycemia [49]. In line with the findings of previous studies in type 2 diabetics, we did not observe any episode of hypoglycemia in NODAT patients treated with vildagliptin after kidney transplantation. Interestingly, preclinical data demonstrate a beneficial effect of gliptins on blood vessels and the heart [50], although a benefit for the cardiovascular system as a whole has still to be proven [27]. Unfortunately, such endpoints are also beyond the scope of the present trial. The same is of course true for statistically meaningful determination of the risk for acute pancreatitis, pancreas dysplasia and thyroid cancer [51-53].

Physical exercise training in KTRs remains a promising tool to affect cardiovascular parameters and outcomes after kidney transplantation, but previous trials are of limited sample size [54, 55]. In this study, lifestyle modification, which included dietary advice and advice on physical activity, did not significantly improve glycemic control, in contrast to a previous study that implemented a structured program [56]. Additional studies are clearly required to explore the efficacy and feasibility of transplant-specific lifestyle intervention programs in patients with NODAT [56].

The present trial has several limitations including the small sample size and the relatively short treatment duration. Inclusion criteria and the primary endpoint were based on OGTT, a method that is regarded as the gold standard for NODAT detection and diagnosis [57]. However, only one OGTT was performed in the present study at baseline as well as 3 and 4 months thereafter. Whether continuous treatment with DPP-4 inhibitors may prove beneficial in NODAT patients will also need to be examined further.

However, we were able to demonstrate that the DPP-4 inhibitor vildagliptin was highly efficient in achieving adequate glycemic control after short-term treatment. Furthermore, vildagliptin was expectably safe regarding hypoglycemic episodes. Adverse events were mild and appeared at similar rates in both study arms. Thus, vildagliptin proved to be a safe and efficient treatment option against NODAT in KTRs, and provides a suitable alternative for this specific form of diabetes.

Acknowledgments

We thank the study volunteers for participating in this investigation and the nursing staff of the Vienna General Hospital for their skilled work. We thank Dr. Josef Kletzmayr for his support in patient recruitment. This academic study was sponsored by the Medical University of Vienna, Austria. Trial registration: ClinicalTrials.gov NCT00980356.

Disclosure

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

Ancillary