SEARCH

SEARCH BY CITATION

Keywords:

  • dipeptidyl peptidase-4;
  • drug-induced liver injury;
  • infections;
  • pancreatitis;
  • renal impairment;
  • skin

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Aim: To assess the safety of vildagliptin versus all comparators (ACs) with regard to organs, systems or tissues of particular interest in type 2 diabetes (T2DM) and areas of potential concern with dipeptidyl peptidase-IV (DPP-4) inhibitors.

Methods: Data were pooled from 38 studies where vildagliptin was given for ≥12 to > 104 weeks in patients with T2DM. Absolute and exposure-adjusted incidence rates and Peto odds ratios (ORs) versus ACs with corresponding 95% confidence intervals (CI) were calculated.

Results: There were > 7000 subject-years of exposure (SYE) to vildagliptin 50 mg bid and > 6500 SYE to ACs. For mild hepatic enzyme elevations with and without elevated bilirubin levels, the ORs for vildagliptin 50 mg bid were 1.24 (95% CI: [0.80, 1.93]) and 1.19 (95% CI: [0.29, 4.90]), respectively. The exposure-adjusted incidences of markedly elevated hepatic enzymes and for enzyme elevations with bilirubin ≥ 2× ULN with vildagliptin 50 mg bid were ≤ those in the ACs group. For hepatic and pancreatitis-related AEs, the ORs for vildagliptin 50 mg bid were 0.87 (95% CI: [0.64, 1.19]) and 0.70 (95% CI: [0.26, 1.88]), respectively, and for any AE in the infections and infestations SOC, this was 1.04 (95% CI: [0.96, 1.13]). The incidences of skin-related AEs were low and the risk with vildagliptin 50 mg bid was not significantly different from ACs [(OR = 1.10 (95% CI: [0.80, 1.51])].

Conclusions: The present meta-analyses indicate that vildagliptin was not associated with increased risk of hepatic events or hepatic enzyme elevations indicative of drug-induced liver injury, pancreatitis, infections or skin-related toxicity.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

The safety of any new therapeutic entity is of great importance, particularly if it is indicated for a chronic and progressive disease such as type 2 diabetes mellitus (T2DM), that requires lifelong treatment. In the case of T2DM, the nature of the disorder itself imposes an increased risk for several organ-specific complications that could be exacerbated by drug treatment. These include drug-induced liver injury (DILI) resulting from high background rate of non-alcoholic fatty liver disease (NAFLD) [1,2] and the association with hepatitis C infection [3,4], pancreatitis [5,6], and a broad spectrum of cardiovascular disease [7,8]. T2DM is also a major risk factor for chronic kidney disease. It was therefore of interest to examine the overall safety of vildagliptin in patients with renal impairment, although at present there is no evidence of a dose-related increase in adverse events (AEs; comparing vildagliptin 50 mg qd to 50 mg bid), and the major pathway for vildagliptin elimination is metabolism (i.e. hydrolysis) [9].

In addition to the aforementioned areas of potential safety concern related to any and all antidiabetic treatments, some specific classes of drugs are associated with additional possible adverse experiences. For the dipeptidyl peptidase-IV (DPP-4) inhibitors, concern has been raised about potential effects on the immune system [10,11]. This concern arises from the fact that the DPP-4 enzyme, also known as the protein CD26, is expressed on activated T cells [12], and sitagliptin treatment was reported to significantly increase the incidence of nasopharyngitis [13]. However, the proteolytic activity of DPP-4/CD26 is not thought to be involved in its role in immune function [14], and DPP-4/CD26 knockout mice appear to be healthy and immunologically competent [15], possibly negating the basis for this mechanism-related concern.

Another potential safety concern specific to DPP-4 inhibitors arises from findings that vildagliptin and some other DPP-4 inhibitors have been associated with dose-dependent necrotic skin lesions when administered to cynomolgus monkeys [16,17] and there have been postmarketing reports of hypersensitivity reactions, including Stevens-Johnson syndrome, and cutaneous vasculitis with the use of sitagliptin [18].

The aim of this work is to critically assess the safety of vildagliptin relative to all comparators (placebo and active comparators pooled) in the areas of concern arising from T2DM itself, as well as those specific to DPP-4 inhibitors. Thus, we report here the results of large meta-analyses of hepatic enzyme elevations and of hepatic-, pancreatitis-, infection- and skin-related AEs, pooling data from Phase II and Phase III clinical studies of vildagliptin, used either as monotherapy or in combination with other oral antidiabetic drugs or insulin. In addition, the general safety of vildagliptin in patients with impaired renal function is evaluated. The cardiovascular safety of vildagliptin is addressed in the companion paper immediately preceding the present report.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Populations

A total of 38 Phase II and Phase III clinical trials that used vildagliptin (50 or 100 mg daily) as monotherapy or in combination with metformin, thiazolidinediones (TZDs), sulfonylureas (SUs) or insulin for ≥ 12 weeks up to > 104 weeks completed as of May 2009 were used for the safety analyses. Table A1 briefly describes each of these studies and indicates which studies were included in each of the four pooled datasets used for the safety analyses.

Because laboratory data are very unlikely to be influenced by the open-label nature of a study, for the categorical analysis of persistent, treatment-emergent transaminase elevations (i.e., those not present at any pretreatment visit and meeting the criterion at consecutive on-treatment measurements or at last on-treatment visit), data from all 38 studies were pooled, and this population is termed the ‘all studies (including open-label) safety population’. Calculation of Peto odds ratios (ORs) requires a comparator, thus calculation of ORs and the resulting Forest plots for hepatic enzyme elevations used data pooled from all controlled trials. This population is termed the ‘all controlled studies (including open-label) safety population’.

For analysis of hepatic- pancreatitis-, immune system- and skin-related AEs, as well as for analysis of AEs and serious AEs (SAEs) in patients with normal renal function and with mild renal impairment, open-label studies were excluded to minimize reporting bias. This population is termed the ‘all studies (excluding open-label) population.’

The population for the calculation of ORs for these AEs used data pooled from all controlled trials excluding open-label trials. This population is termed the ‘all controlled studies (excluding open-label) population’.

Assessments

All AEs were recorded and assessed by the investigator as to the severity and possible relationship to the study medication. All laboratory assessments were performed by central laboratories: Bioanalytical Research Corporation-USA (Lake Success, NY, USA), Bioanalytical Research Corporation-EU (Ghent, Belgium), Diabetes Diagnostics Laboratory (Columbia, MO, USA), Covance (Geneva, Switzerland, Singapore, or Indianapolis, IN, USA) or Medical Research Laboratories International (Zaventem, Belgium).

Standardization of Terms

AEs were encoded in all studies using the Medical Dictionary for Regulatory Activities (MedDRA, version 12.0) system. This is a medically validated terminology database developed by the International Conference on Harmonisation. Within the MedDRA, AEs are grouped by type of AE, such as ‘infections and infestations' or by body system, for example ‘cardiac disorders', with groupings referred to as ‘system organ classes' (SOCs). Within an SOC, specific AEs are identified by ‘preferred term’ (PT). The PTs included in this analysis of selected hepatic-, pancreatitis- and skin-related AEs are listed in Table A2. The PTs used for analysis of skin-related safety were chosen to provide the closest clinical correlation to the skin lesions observed in preclinical safety studies in cynomolgus monkeys. For analysis of potential effects on the immune system, all PTs in the SOC of infections and infestations were included.

Data Analysis

For each analysis population, we provide the total number of patients and subject-years of exposure [SYE]. Pooled demography and baseline characteristics in the all studies (excluding open-label) safety population and the incidence of selected AEs in absolute numbers and adjusted for exposure are also presented.

For each trial Peto OR and corresponding 95% confidence intervals (CIs) were calculated to compare selected AEs between vildagliptin and the all comparator groups. Pooled estimates were obtained using a fixed effect model and presented in forest plots. OR below unity were indicative of a treatment effect favouring vildagliptin. Correction for continuity using the inverse of the opposite arm size was used when zero events occurred. This correction causes less bias than the standard continuity correction of 0.5 [19] when the size of the treatment arms are unbalanced [20]. Between-trial heterogeneity was assessed by Cochrane's Q-test, using p ≤ 0.10 as indicative of significant heterogeneity, and by the scale-free I2 index, representing the percent of variability between trials attributable to heterogeneity rather than chance [21].

Ethics and Good Clinical Practice

All study participants provided written informed consent. All protocols were approved by the independent ethics committee/institutional review board at each study site or country. All studies were conducted using good clinical practice and in accordance with the Declaration of Helsinki.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Exposure

The number of patients and the total SYE for vildagliptin 50 mg qd, vildagliptin 50 mg bid and for all comparators (pooled) are summarized for each safety population in Table A3. In the all studies excluding open-label safety population (used for categorical analyses of AEs), 2049 patients received vildagliptin 50 mg qd, 6116 patients received vildagliptin 50 mg bid and 6210 patients received any comparator. Because the number of patients exposed to vildagliptin was substantially higher for the 50 mg bid group than the 50 mg qd group and because any potential risk would be expected to be manifest at the higher dose, data from this group are the main focus of this work.

Demography

Table 1 summarizes baseline characteristics of patients in the all studies (excluding open-label) safety population. The demographic and baseline characteristics of patients were comparable across the three treatment groups. The mean age, BMI, HbA1c and fasting plasma glucose (FPG) were approximately 56 years, 31.4 kg/m2, 8.1% and 9.8 mmol/l, respectively. Patients on average had been diagnosed with T2DM more than 4 years previously, and nearly one third of patients had some degree of renal insufficiency. Demographic and background characteristics of the other populations were generally similar and overall, the populations were representative of a full spectrum of patients with T2DM.

Table 1.  Demography and baseline characteristics [all studies (excluding open-label) safety population].
Mean ± s.d. or N (%)Vilda 50 mg qd N = 2049Vilda 50 mg bid N = 6116All comparators N = 6210
  1. GFR (MDRD), glomerular filtration rate calculated by the Modification of Diet in Renal Disease equation.

Age (years)55.6 ± 10.955.4 ± 10.556.5 ± 10.6
 Age group (years)   
 < 651614 (78.8)4838 (79.1)4623 (74.4)
 ≥ 65435 (21.2)1278 (20.9)1587 (25.6)
Sex   
 Male1131 (55.2)3363 (55.0)3399 (54.7)
 Female918 (44.8)2753 (45.0)2811 (45.3)
Race   
 Caucasian1473 (71.9)4458 (72.9)4514 (72.7)
 Hispanic or Latino231 (11.3)660 (10.8)774 (12.5)
 Black150 (7.3)235 (3.8)224 (3.6)
 Asian (non-Indian subcontinent)82 (4.0)487 (8.0)403 (6.5)
 Asian (Indian subcontinent)78 (3.8)185 (3.0)161 (2.6)
 All other35 (1.7)91 (1.5)134 (2.1)
BMI (kg/m2)31.6 ± 5.531.4 ± 5.431.3 ± 5.4
 BMI group (kg/m2)   
 < 30897 (43.8)2692 (44.0)2840 (45.7)
 ≥ 301150 (56.1)3413 (55.8)3362 (54.1)
 Not recorded2 (0.1)11 (0.2)8 (0.1)
HbA1c (%)8.1 ± 1.38.3 ± 1.18.0 ± 1.1
 HbA1c group (%)   
 ≤ 81037 (50.6)3002 (49.1)3506 (56.5)
 > 8931 (45.4)3037 (49.7)2621 (42.2)
 Not recorded81 (4.0)77 (1.3)83 (1.3)
FPG (mmol/l)9.3 ± 2.810.1 ± 2.79.7 ± 2.7
Duration of T2DM (years)4.5 ± 5.64.2 ± 5.14.6 ± 5.2
GFR (MDRD) (ml/min) per (1.73 m2)   
 Normal (> 80)1338 (65.3)4232 (69.2)4217 (67.9)
 Mild (≥ 50, ≤ 80)665 (32.5)1802 (29.5)1918 (30.9)
 Moderate (≥ 30, < 50)38 (1.9)76 (1.2)74 (1.2)
 Severe (< 30)4 (0.2)5 (0.1)1 (0.0)
 Not recorded4 (0.2)1 (0.0)0 (0.0)

Hepatic Safety

Hepatic safety was assessed by examining both persistent treatment-emergent liver enzyme elevations and hepatic AEs and SAEs. The Peto OR of having an alanine aminotransferase and/or aspartate aminotransferase (ALT/AST) ≥ 3 times the upper limit of normal (ULN) was slightly, but not significantly (95% CI included 1.0) above unity for both vildagliptin dose regimens verus all comparators (Figure 1). The Peto OR estimate for vildagliptin 50 mg qd versus all comparators was greater than that for vildagliptin 50 mg bid and was associated with broader 95% CI, reflecting the considerably smaller number of patients receiving the lower dose. Overall, there was no evidence of an increased risk for ALT/AST ≥ 3× ULN accompanied by bilirubin > ULN (an indicator of more severe hepatic injury); the ORs relative to all comparators were 0.21 for vildagliptin 50 mg qd and 1.19 for vildagliptin 50 mg bid, with no statistically significant differences (95% CIs included 1.0). Of note, these OR estimates are based on a low number of events, explaining the fairly wide associated CIs. Statistical measures of heterogeneity (Q-test and I2 index) showed consistency of effect across studies contributing to these results (Q between 0.23 to 9.59, p-value between 0.79 and 0.99 and I2 = 0.00).

image

Figure 1. Odds ratios for treatment-emergent persistent on-treatment hepatic enzyme elevations in the all controlled studies (including open-label) safety population and selected hepatic AEs and SAEs by treatment in the all controlled studies (excluding open-label) safety population. (Vilda = vildagliptin; All comparators = all non-Vilda treatment groups, ie, placebo and active comparators. n = number of patients meeting the criterion or experiencing an event; N = total number of patients with evaluable criterion or total number of patients. Persistent elevations are those which met the criterion at consecutive on-treatment measurements or at last on-treatment measurement). Test for heterogeneity of alanine aminotransferase and/or aspartate aminotransferase (ALT/AST) ≥ 3× upper limit of normal (ULN): Q = 2.73, p = 0.997 and I2 = 0.00 (vildagliptin 50 mg qd); Q = 9.59, p = 0.792 and I2 = 0.00 (vildagliptin 50 mg bid). Test for heterogeneity of ALT/AST ≥ 3× ULN and bilirubin > ULN: Q = 0.23, p = 0.999 and I2 = 0.00 (vildagliptin 50 mg qd); Q = 2.34, p = 0.999 and I2 = 0.00 (vildagliptin 50 mg bid). Test for heterogeneity of selected hepatic AEs: Q = 8.02, p = 0.784 and I2 = 0.00 (vildagliptin 50 mg qd); Q = 17.10, p = 0.251 and I2 = 0.18 (vildagliptin 50 mg bid). Test for heterogeneity of selected hepatic SAEs: Q = 2.92, p = 0.996 and I2 = 0.00 (vildagliptin 50 mg qd); Q = 5.02, p = 0.986 and I2 = 0.00 (vildagliptin 50 mg bid).

Download figure to PowerPoint

Kaplan Meier analysis of treatment-emergent persistent ALT/AST ≥ 3× ULN showed that there was no specific pattern of onset of events for the highest vildagliptin dose of 50 mg bid and all comparators (data not shown).

The incidences of markedly elevated persistent hepatic enzyme elevations (ALT/AST ≥ 10× or ≥ 20× ULN) were very low across all treatment groups and similar in the vildagliptin 50 mg bid group (1/5917 patients for ALT/AST ≥10× ULN and 0/5917 patients for ALT/AST ≥ 20× ULN), and the all comparators group (2/6695 patients for ALT/AST ≥ 10× ULN and 0/6695 patients for ALT/AST ≥ 20× ULN). This was also the case for ALT or AST ≥ 3× ULN accompanied by bilirubin ≥ 2× ULN; that is the incidences were again very low across treatment groups with a similar incidence for vildagliptin 50 mg bid (3/5906 patients) and all comparators (3/6595 patients). There were no such events with vildagliptin 50 mg qd, and the exposure-adjusted incidences were identically low for vildagliptin 50 mg bid and all comparators (Table 2). Two of the three cases of ALT or AST ≥ 3× ULN accompanied by bilirubin ≥ 2× ULN on vildagliptin 50 mg bid were counfounded by alternative etiologies (metastatic pancreatic adenocarcinoma with common bile duct stricture, and history of hepatic steatosis with elevated ALT, AST and ALP prior to start of study medication, respectively). The other case was characterized by abnormal liver enzyme tests only, clinically asymptomatic and with complete resolution when vildagliptin treatment was terminated.

Table 2.  Absolute and exposure-adjusted incidences of treatment-emergent persistent on-treatment hepatic enzyme elevations (all studies [including open-label] safety population).
  Vildagliptin50 mg qd N = 2206Vildagliptin 50 mg bid N = 6159All comparators N = 7082
  1. Total = Number of patients with evaluable criterion. n = Number of patients meeting the criterion. Notable abnormalities summarized are those which are treatment-emergent (i.e. not present at any pretreatment visit) and persistent (meeting criterion at consecutive on-treatment measurements or at last on-treatment visit). SYE-adj = SYE-adjusted = incidence per 100 SYE defined as 100 x (number of patients with any event divided by the total exposure time in years).

ALT or ASTTotal209159176695
 ≥ 10 × ULNn (%)0 (0.0)1 (0.0)2 (0.0)
 SYE-adj0.000.010.03
 ≥ 20 × ULNn (%)0 (0.0)0 (0.0)0 (0.0)
 SYE-adj0.000.000.00
ALT or AST ≥ 3 × ULN and bilirubin ≥ 2 × ULNTotal208559066595
 n (%)0 (0.0)3 (0.1)3 (0.0)
 SYE-adj0.000.040.04

There was also no evidence of increased risk for hepatic AEs or SAEs with vildagliptin treatment. The Peto OR for vildagliptin 50 mg bid relative to all comparators for selected hepatic AEs was < 1, with relatively narrow 95% CIs, and for selected hepatic SAEs the Peto OR was slightly >1, but with broader 95% CIs, that included 1.0 (Figure 1). No heterogeneity was found (Q p-value > 0.25, I2≤ 0.18). For either vildagliptin dose regimen, the hepatic AEs were generally considered to be mild; they were transient in nature and seldom required clinical intervention or drug discontinuation.

The incidences and exposure-adjusted incidences of any selected hepatic AE or SAE were lower with vildagliptin 50 mg bid than for all comparators (Table 3). The most common specific hepatic AEs were hepatic steatosis and ALT increased, and for these AEs the incidences were also lower with vildagliptin 50 mg bid (0.5%; 0.45 events per 100 SYE and 0.2%; 0.16 events per 100 SYE, respectively) than for all comparators (0.6%; 0.54 events per 100 SYE and 0.3%; 0.26 events per 100 SYE, respectively). For all other specific hepatic AEs, the incidences with vildagliptin 50 mg bid were similar to or lower than with all comparators. The incidences of any specific hepatic SAE were limited to ≤ 2 patients in any treatment group, and the incidences and exposure-adjusted incidences of any specific hepatic SAE were similar with vildagliptin 50 mg bid and all comparators. The one SAE of acute hepatic failure occurred in a patient receiving placebo.

Table 3.  Absolute and exposure-adjusted incidences of selected hepatic adverse events (AEs) and serious AEs (SAEs), pancreatitis-related AEs, AEs in the infections and infestations System Organ Class (SOC) and selected skin-related AEs and SAEs. [All studies (excluding open-label) safety population].
Event categoryVilda 50 mg qdVilda 50 mg bidAll comparators
N = 2049 SYE = 1253.9N = 6116 SYE = 7313.6N = 6210 SYE = 6512.7
n(%)n(%)n(%)
SYE adjSYE-adjSYE-adj
  1. SYE-adjusted (SYE-adj) rates calculated as number of patients having events per 100 subject year exposure.

Hepatic
 Any hepatic AE26 (1.3)85 (1.4)93 (1.5)
 2.071.161.43
 Any hepatic SAE2 (0.1)7 (0.1)8 (0.1)
 0.160.100.12
Pancreatitis
 Any pancreatitis-related AE3 (0.1)7 (0.1)11 (0.2)
 0.240.100.17
Immune system
 Infections and Infestations SOC; any AE522 (25.5)2162 (35.3)2014 (32.4)
 41.6329.5630.92
Skin
 Any skin-related AE26 (1.3)95 (1.6)78 (1.3)
 2.071.301.20
 Any skin-related SAE1 (0.0)6 (0.1)7 (0.1)
 0.080.080.11

Pancreatitis

As depicted in Figure 2, there was no evidence of an increased risk of pancreatitis-related AEs with vildagliptin. The Peto OR for vildagliptin 50 mg bid versus all comparators was < 1, as was the OR for vildagliptin 50 mg qd; both had relatively broad 95% CIs, reflecting a low total number of patients. No heterogeneity was found (Q p-value > 0.98, I2 = 0.00). The incidences and SYE-adjusted incidences of any pancreatitis-related AE (Table 3) and of any specific pancreatitis-related AE were very low for each pooled treatment group and were lower with vildagliptin 50 mg bid than with all comparators.

image

Figure 2. Odds ratios for pancreatitis-related adverse events (AEs) and AEs in the Infections/Infestations System Organ Class (SOC) in the all controlled studies (excluding open-label) safety population. (Vilda = vildagliptin; All comparators = all non-Vilda treatment groups, that is placebo and active comparators. n = number of patients experiencing an AE upper limit of normal, N = total number of patients ). Test for heterogeneity of selected pancreatitis-related AEs: Q = 3.57, p = 0.990 and I2 = 0.00 (vildagliptin 50 mg qd); Q = 4.51, p = 0.992 and I2 = 0.00 (vildagliptin 50 mg bid). Test for heterogeneity of AEs in the ‘Infections and Infestations' SOC: Q = 9.51, p = 0.659 and I2 = 0.00 (vildagliptin 50 mg qd); Q = 20.0, p = 0.130 and I2 = 30.04 (vildagliptin 50 mg bid).

Download figure to PowerPoint

Immune System

There was also no evidence of increased risk of immune system suppression with vildagliptin therapy. The Peto ORs for AEs in the SOC of infections and infestations with both vildagliptin dose regimens relative to all comparators were very close to 1, with narrow 95% CIs (Figure 2). Again, no heterogeneity was found for these analyses (Q p-value > 0.12, I2≤ 30.04). The overall exposure-adjusted incidences of AEs in the infections and infestations SOC were comparable between vildagliptin 50 mg bid (29.6/100 SYE) and all comparators (30.9/100 SYE) (Table 3), as were the incidences of any specific AE. The incidences of immune system-related SAEs (≤ 1.5%) and discontinuations because of AEs (≤ 0.3%) were low and similar across treatment groups.

Skin

The risk for selected skin-related AEs for vildagliptin 50 mg bid was similar to all comparators with the OR being only slightly above 1, and with the 95% CIs including 1.0, indicating no statistical significance. The Peto OR for vildagliptin 50 mg qd versus all comparators was higher, but still not statistically significant, and with broader 95% CI, reflecting the much smaller sample size (Figure 3). Homogeneity of the trial estimates for both analyses was confirmed (Q p-value > 0.65, I2 = 0.00). The Peto ORs for selected skin-related SAEs with both vildagliptin treatment regimens versus all comparators were <1, indicating no increased risk with vildagliptin (Figure 3). Again, we observed homogeneity in the estimates for the analysed trials (Q p-value > 0.72, I2 = 0.00).

image

Figure 3. Odds ratios for selected skin and vascular-related AEs and SAEs in the all controlled studies (excluding open-label) safety population. (Vilda = vildagliptin; All comparators = all non-Vilda treatment groups, that is placebo and active comparators. n = number of patients experiencing an AE , N = total number of patients .) Test for heterogeneity of selected skin- and/or vascular-related AEs: Q = 9.58, p = 0.653 and I2 = 0.00 (vildagliptin 50 mg qd); Q = 10.79, p = 0.702 and I2 = 0.00 (vildagliptin 50 mg bid). Test for heterogeneity of selected skin- and/or vascular-related SAEs: Q = 0.20, p = 0.999 and I2 = 0.00 (vildagliptin 50 mg qd); Q = 10.31, p = 0.739 and I2 = 0.00 (vildagliptin 50 mg bid).

Download figure to PowerPoint

The overall incidence of skin-related AEs was low across all treatment groups. The SYE-adjusted incidence of any skin-related AE in patients receiving vildagliptin 50 mg bid was similar to all comparators (Table 3). The most commonly reported AEs were skin ulcer, blister and skin lesion, although the incidences were low. The SYE-adjusted rates for skin ulcer and blister were similar for vildagliptin 50 mg bid (0.27 events per 100 SYE and 0.18 events per 100 SYE, respectively) and all comparators (0.26 events per 100 SYE and 0.21 events per 100 SYE, respectively), whereas the incidence of ‘skin lesion’ was slightly higher with vildagliptin 50 mg bid (0.27 events per 100 SYE) than with all comparators (0.14 events per 100 SYE). No patient discontinued because of a skin lesion AE with vildagliptin, whereas there was one discontinuation in the all comparators group. The overall incidence of skin-related SAEs was very low, and the SYE-adjusted incidences of any skin-related SAE were somewhat lower for either vildagliptin dose regimen than for all comparators (Table 3).

Safety in Patients with Impaired Renal Function

More than 30% of patients exposed to vildagliptin 50 mg qd or bid in the all studies (excluding open-label) safety database had mild renal impairment, thus exposure to vildagliptin in this group was sufficient to allow a meaningful assessment of safety. In contrast, the number of patients in the safety database with moderate or severe renal impairment was insufficient to support such an analysis (cf. Table 1). Table 4 summarizes incidences of AEs and SAEs in patients with normal and mildly impaired renal function. In patients with normal renal function, the incidences of any AE or any SAE were similar in the vildagliptin 50 mg bid group and the all comparators group. Although the incidence of AEs across treatment groups in patients with mild renal impairment was slightly higher than in patients with normal renal function, as would be expected in a renally impaired population, the incidences of any AE or any SAE in patients receiving vildagliptin 50 mg bid were again similar to those in the all comparators group. Thus, the presence of mild renal impairment did not adversely affect the safety of vildagliptin relative to patients with normal renal function.

Table 4.  Incidences of AEs and SAEs in patients with normal renal function and mild renal impairment (all studies [excluding open-label] safety population).
  Vilda 50 mg qdVilda 50 mg bidAll comparators
Normal renal function: GFR (MDRD) > 80 (ml/min) × (1.73 m2)
 Patient number (N)  133842324217
 Any AEn (%)786 (58.7)2895 (68.4)2897 (68.7)
 Any SAEn (%)56 (4.2)316 (7.5)332 (7.9)
Mild renal impairment: GFR (MDRD) ≥ 50 but ≤ 80 (ml/min) × (1.73 m2)
 Patient number (N)  66518021918
 Any AEn (%)436 (65.6)1267 (70.3)1345 (70.1)
 Any SAEn (%)54 (8.1)209 (11.6)214 (11.2)

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

The present analyses of data pooled from up to 38 Phase II and Phase III clinical trials with vildagliptin were undertaken to address several specific safety topics of concern in patients with T2DM and/or of potential concern with DPP-4 inhibitors. For the meta-analyses presented no significant heterogeneity was found across trials. Overall, the data from moer than 8000 patients treated with vildagliptin (50 mg qd or bid) in studies of up to more than 2 years in duration raise no safety signals related to severe DILI, pancreatitis, or immune system function (infections and infestations), and the human data reported herein are not consistent with the necrotic skin lesions seen with vildagliptin and some other DPP-4 inhibitors in toxicology studies in monkeys.

Because patients with T2DM have a high incidence of NAFLD, and several antidiabetic agents such as acarbose, gliclazide and metformin have been associated with DILI [22], the hepatic safety of any new diabetes treatment is of interest. This is also relevant because of the putative association between T2DM and hepatitis C infection [3,4] another risk factor potentially confounding the hepatic safety evaluation of antidiabetic drugs. In this pooled hepatic safety analysis, vildagliptin was associated with a slightly increased OR of mild persistent transaminase elevations (≥ 3× ULN) relative to all comparators, which was not statistically significant. Overall, there was no increased risk of persistent transaminase elevations with concomitant bilirubin elevation > ULN for either vildagliptin dose. Importantly, the incidences of persistent ALT/AST elevations of ≥ 10× ULN, ≥20× ULN, or ≥ 3× ULN with concomitant bilirubin elevation ≥ 2× ULN were very low with vildagliptin and similar to all comparators. Hence, the risk of transaminase elevations considered to be more specific for potential drug-induced hepatotoxicity is low and similar between vildagliptin and all comparators.

It is worth noting that the incidence of transaminase elevations ≥ 3× ULN observed with viladgliptin is in the range to that reported for the widely used drug, simvastatin, for which persistent increases (to more than 3× the ULN) in serum transaminases have occurred in approximately 1% of patients who received the drug in clinical studies [23].

The incidence of hepatic AEs with vildagliptin was low, and when reported, they were mild and transient in nature, seldom requiring clinical intervention or drug discontinuation [24]. With vildagliptin treatment, hepatic SAEs were rare, with no apparent relationship to dose and they occurred at a rate similar to that seen with all comparators.

To date, there is little evidence that vildagliptin or other DPP4 inhibitors are associated with significant hepatic risk. Although cases of ALT elevations with concomitant increase in bilrubin have been reported recently for sitagliptin, these cases resolved on treatment and overall no increased risk of hepatic events was reported [25]. Given that a small number of these hepatic enzyme elevations were reported on vildagliptin as well, liver enzyme monitoring after initiation of therapy is prudent and consistent with the vildagliptin product information.

T2DM has also been associated with increased risk of pancreatitis compared to the general population as several clinical factors associated with T2DM and obesity (i.e. cholelithiasis, hypertrygliceridaemia) are known to be putative risk factors for acute pancreatitis [5,26]. GLP-1 agonists such as exenatide have been associated with some cases of acute pancreatitis [27], as has sitagliptin [28]; accordingly, pancreatitis-related AEs were also assessed in vildagliptin-treated patients. The incidence of any pancreatitis-related AE was very low across all treatment groups and the ORs for both vildagliptin 50 mg qd and bid were less than 1.0; the exposure-adjusted incidence with vildagliptin 50 mg bid (0.10 patients per 100 SYE) was slightly less than that with all comparators (0.17 patients per 100 SYE). Thus, there is no apparent increased risk of pancreatitis with the DPP-4 inhibitor vildagliptin. Furthermore, there is no mechanistic basis for this toxicity and there is no evidence for a causal link between pharmacological DPP4 inhibition and pancreatitis.

The potential for impaired immune function is an issue that has been raised for all DPP-4 inhibitors [10] because the enzyme DPP-4 is also a protein (CD26) expressed on activated T lymphocytes [12], and CD26 is thought to be important for many aspects of immune function [29,30]. However, this analysis of AEs in the SOC of infections and infestations yielded ORs for either vildagliptin dose regimen very close to 1.0, with narrow CIs, and suggests that vildagliptin does not increase risk of infection. Furthermore, the exposure-adjusted incidences of commonly reported specific AEs such as nasopharyngitis and upper respiratory tract infections were very similar to vildagliptin and all comparators. This is consistent with a recent report describing a pooled safety analysis of sitagliptin in which a slightly increased risk of AEs in this SOC relative to comparators was not statistically significant [25]. The lack of a safety signal for infections with DPP-4 inhibitors may be because that the enzymatic (peptidase) activity of the DPP-4/CD26 protein is not thought to be involved in immune function [14].

In addition to the clinical data suggesting no significant increase in infections and infestations (i.e. no suppression of immune function) with vildagliptin treatment reported herein, preclinical data also provide no evidence of immune suppression during vildagliptin administration. Thus, daily administration of a wide range of doses vildagliptin for 4 weeks in rats suppressed neither the primary nor the secondary response [increased immunoglobulin G (IgG) and immunoglobulin M (IgM) titres] to immunization with keyhole limpet hemocyanin. Furthermore, vildagliptin administration did not affect the development of immunization-related histological (splenic, lymph nodes) and morphological changes (injection site granuloma formation) [31].

Additional concerns specific to DPP-4 inhibitors arise from the fact that the enzyme DPP-4 has many possible substrates other than the incretin hormones [32]. Bradykinin and substance P are among these possible substrates. Inhibition of the degradation of bradykinin and substance P has been suggested to be responsible for an increased rate of angioedema associated with use of angiotensin-converting enzyme inhibitors (ACEi) and DPP-4 activity was reported to be significantly reduced in patients with a history of ACEi-associated angioedema [33]. It was recently reported that in a prospective, adjudicated analysis of angioedema-related AEs, although vildagliptin treatment was not associated with increased risk of angioedema, concomitant therapy with an ACEi and vildagliptin was associated with an increased risk of angioedema. This analysis, however, was limited by the relatively low number of cases and most cases of angioedema in patients receiving vildagliptin and an ACEi were of mild or moderate clinical severity, with the majority of events resolving despite continued use of vildagliptin [34].

In addition to enzyme specificity issues discussed above, it has been suggested that lack of absolute inhibitor selectivity for DPP-4 may give rise to unwanted side effects, or toxicity. Thus, some preclinical studies using various selective and non-selective inhibitors of DPP-4 and related peptidases have concluded that the lack of selectivity of some DPP-4 inhibitors relative to DPP-8 and/or DPP-9 leads to multiorgan toxicities in rodents [35,36]. However, we have recently provided data from preclinical studies with vildagliptin which refute this hypothesis [37], showing that none of the toxicities reported in rodents with a non-selective inhibitor were reproduced during studies with very high doses of vildagliptin that were sufficient to completely inhibit both DPP-8 and DPP-9.

Another safety concern specific to DPP-4 inhibitors is that of possible skin-related AEs. Vildagliptin, saxagliptin and dutogliptin (PHX-1149), have been found to elicit dose-dependent necrotic skin lesions in monkeys [17,25], and the spectre of DPP-8/9 inhibition was raised as a possible explanation. Therefore, it was also of interest to carefully assess skin-related AEs in this work; however, there was no evidence of a skin-related safety signal with vildagliptin treatment. Although the lower dose of vildagliptin was associated with an OR > 1.0 (albeit non-significant) for skin-related AEs, the OR for the higher dose treatment regimen was close to 1. The latter analysis was based on much larger patient numbers and is therefore considered more robust. Furthermore, for skin-related SAEs the ORs were < 1.0 indicating lack of increased risk for clinically relevant skin findings relative to comparators.

Because skin-related toxicities have not been seen with vildagliptin or reported with other DPP4 inhibitors, such as saxagliptin in species other than the monkey, skin-related toxicity is considered a species-specific finding in monkeys.

The present assessment of vildagliptin safety also included determination of incidence rates of any AE and any SAE in patients with mild renal impairment. It was found that the incidences of both AEs and SAEs were very similar with vildagliptin 50 mg bid and all comparators. Thus, mild renal impairment does not appear to raise any safety signal with vildagliptin relative to patients with normal renal function. Two dedicated, large studies are currently ongoing to assess the long-term safety of vildagliptin in patients with moderate and severe renal impairment.

In summary, the present targeted meta-analyses of data from more than 8000 vildagliptin-treated patients was undertaken to address several specific safety topics that are of particular concern for patients with T2DM and others that are specific for DPP-4 inhibitors. The data suggest that vildagliptin does not increase risk of hepatic events or hepatic enzyme elevations indicative of DILI, pancreatitis-, immune system- or skin-related AEs in patients with T2DM. Furthermore, there were no safety signals in patients with T2DM and mild renal impairment, and as addressed in the companion paper, there was no evidence of increased cardiovascular risk with vildagliptin. Because T2DM is a chronic and progressive disease that requires long-term treatment, it is of critical importance to consider all potential risks relative to the potential benefit of a therapeutic agent or treatment regimen. Vildagliptin and other DPP-4 inhibitors leverage a novel, physiological mechanism and have been shown to be an effective approach to diabetes treatment [11,13,38,39]. The present data including studies of up to more than 2 years duration should reduce concern about several specific risks, but continued surveillance is required to ultimately determine the risk/benefit balance for vildagliptin and other incretin-based therapeutics.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

The authors acknowledge the patients, investigators, and staff at participating sites for all the studies and Beth Dunning Lower for editorial assistance and helpful discussion. This work was funded by Novartis Pharmaceuticals Corporation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Appendix

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix
Table A1.  Vildagliptin studies contributing to meta-analyses.
Study No.Study descriptionPhase/populationRandomized patients*Treatment duration**Publication***
  1. *for extension studies: patients who entered extension.

  2. **for extension studies: duration of core + extension study

  3. ***ClinicalTrials.gov Identifier number is provided if data are not yet published

  4. IGT, impaired glucose tolerance; SU, sulfonylurea; T2DM, type 2 diabetes; TZD, thiazolidinedione.

  5. Population a = all studies (including open-label) safety population.

  6. Population b = all controlled studies (including open-label) safety population.

  7. Population c = all studies (excluding open-label) safety population.

  8. Population d = all controlled studies (excluding open-label) safety population.

Monotherapy
1Dose-ranging study in drug-naïve T2DM patients (HbA1c 6.8–10%)II/a,b,c,d27912 weeks[40]
2Uncontrolled 40-week extension to Study 1II/a,c14152 weeksNot available
3Placebo-controlled low-dose efficacy/safety study in drug-naïve T2DM patients (HbA1c: 6.8–11%)II/a,b,c,d10012 weeks[41]
4Placebo-controlled dose-ranging study (efficacy/safety) in drug-naïve T2DM patients (HbA1c 7.5–10%)III/a,b,c,d63224 weeks[42]
5Uncontrolled 28-week extension to Study 4III/a,c44052 weeksNCT00138541
6Placebo-controlled long-term efficacy/safety study in drug-naïve T2DM patients with mild hyperglycemia (HbA1c 6.2–7.5%)III/a,b,c,d30652 weeks[43,44]
7Placebo-controlled 52-week extension to Study 6III/a,b,c,d131104 weeks[45]
8Active-controlled (metformin) long-term efficacy/safety study in drug-naïve T2DM patients (HbA1c 7.5–11%)III/a,b,c,d78052 weeks[46]
9Active-controlled (metformin) 52-week extension to Study 8III/a,b,c,d463104 weeks[47]
10Active-controlled (gliclazide) long-term efficacy/safety study in drug-naïve T2DM patients (HbA1c 7.5–11%)III/a,b,c,d1092104 weeks[48]
11Active-controlled (acarbose) efficacy/safety study in drug-naïve T2DM patients (HbA1c 7.5–11%)III/a,b,c,d66124 weeks[49]
12Active-controlled (rosiglitazone) efficacy/safety study in drug-naïve T2DM patients (HbA1c 7.5–11%)III/a,b,c,d78624 weeks[50]
13Active-controlled (rosiglitazone) 80-week extension to Study 12III/a,b,c,d598104 weeks[51]
14Active-controlled (pioglitazone) dose regimen comparison study in drug-naïve T2DM patients (HbA1c 9–11%)III/a,b,c,d27312 weeksNCT00101673
15Placebo-controlled efficacy/safety study in patients with IGTIII/a,b,c,d17912 weeks[52]
16Placebo-controlled mechanistic study (beta-cell function) in drug-naïve T2DM patients with mild hyperglycemia (HbA1c≤ 7.5%)III/a,b,c,d8952 weeksNCT00260156
17Placebo-controlled dose-ranging study (efficacy/safety) in drug-naïve T2DM patients (HbA1c 7.5–10%)III/a,b,c,d35424 weeks[53]
18Active-controlled (metformin) efficacy/safety study in drug-naïve elderly (≥ 65 years) T2DM patients (HbA1c 7–9%)III/a,b,c,d33524 weeks[54]
Combination therapy with metformin
19Placebo-controlled dose-selection study in patients inadequately controlled by metformin (HbA1c 7.0–9.5%)II/a,b,c,d13212 weeks[55]
20Placebo-controlled 40-week extension to Study 19II/a,b,c,d7152 weeks[55]
21Placebo-controlled efficacy/safety study in T2DM patients inadequately controlled with metformin (HbA1c 7.5–11%)III/a,b,c,d54424 weeks[56]
22Uncontrolled 28-week extension to Study 21III/a,c41752 weeksNCT00138515
23Active-controlled (glimepiride) long-term efficacy/safety study in T2DM patients treated with metformin (HbA1c >6.5–8.5%)III/a,b,c,d3118≥104 weeks[55] NCT00106340
24Active-controlled (gliclazide) long-term efficacy/safety study in T2DM patients inadequately controlled with metformin (HbA1c 7.5–11%)III/a,b,c,d100752 weeksNCT00102466
25Active-controlled (pioglitazone) long-term efficacy/safety study in T2DM patients inadequately controlled with metformin (HbA1c 7.5–11%)III/a,b,c,d57652 weeks[57,58]
26Placebo-controlled efficacy/safety in T2DM patients inadequately controlled with metformin (HbA1c 7.5–11%) to compare a.m. versus p.m. dosing regimensIII/a,b,c,d37024 weeks[59]
27Efficacy/safety study in T2DM patients treated with metformin (HbA1c 6.5–9%) to compare vildagliptin as add-on to metformin versus up-titration of metforminIII/a,b,c,d91424 weeksNCT00396357
28Active-controlled, open-label efficacy/safety study in T2DM patients treated with metformin (HbA1c 7.0–10.0%) in a community-based practice settingIII/a, b266412 weeks[60]
29Efficacy/safety study of initial fixed combination therapy of vildagliptin and metformin in drug-naïve T2DM patients (HbA1c 7.5–11%)III/a,b,c,d117924 weeks[61]
30Open-label substudy of study 29 in drug-naive T2DM patients with very poor glycemic control (HbA1c >11.0%)III/a, b9424 weeksNCT00468039
Combination therapy with TZD
31Placebo-controlled efficacy/safety study in T2DM patients inadequately controlled by TZD (HbA1c 7.5–11%)III/a,b,c,d46324 weeks[50]
32Uncontrolled 28-week extension to Study 31III/a,c31252 weeksNCT00138554
33Uncontrolled, open-label 40-week extension to Study 14III/a20052 weeksNCT00138593
34Initial combination (vildagliptin/pioglitazone) efficacy/safety study in drug-naïve T2DM patients (HbA1c 7.5–11%)III/a,b,c,d60724 weeks[62]
Combination therapy with SU
35Placebo-controlled efficacy/safety study in T2DM patients inadequately controlled by SU (HbA1c 7.5–11%)III/a,b,c,d51524 weeks[63]
36Uncontrolled 28-week extension to Study 35III/a,c33252 weeksNCT00138580
Combination therapy with insulin
37Placebo-controlled efficacy/safety study in T2DM patients treated with insulin (HbA1c 7.5–11%)III/a,b,c,d29624 weeks[64,65]
38Uncontrolled 28-week extension to Study 37III/a,c20052 weeks[65]
Table A2.  Preferred terms included in selected hepatic, pancreatitis and skin-related AEs.
Selected hepatic AEs
5′ nucleotidase increasedHepatic sequestration
Acute graft versus host disease in liverHepatic steatosis
Acute hepatic failureHepatic vascular resistance increased
Alanine aminotransferase abnormalHepatitis
Alanine aminotransferase increasedHepatitis acute
Ammonia abnormalHepatitis cholestatic
Ammonia increasedHepatitis chronic active
Antithrombin III decreasedHepatitis chronic persistent
AscitesHepatitis fulminant
Aspartate aminotransferase abnormalHepatobiliary disease
Aspartate aminotransferase increasedHepatobiliary scan abnormal
AsterixisHepatocellular foamy cell syndrome
Autoimmune hepatitisHepatocellular injury
BacterascitesHepatomegaly
Bile output abnormalHepatopulmonary syndrome
Bile output decreasedHepatorenal failure
Biliary cirrhosisHepatorenal syndrome
Biliary cirrhosis primaryHepatosplenomegaly
Biliary fibrosisHepatotoxicity
Bilirubin conjugated abnormalHyperammonaemia
Bilirubin conjugated increasedHyperbilirubinaemia
Bilirubin excretion disorderHypercholia
Biopsy liver abnormalHypertransaminasaemia
Blood alkaline phosphatase abnormalHypoalbuminaemia
Blood alkaline phosphatase increasedIcterus index increased
Blood bilirubin abnormalInternational normalised ratio abnormal
Blood bilirubin increasedInternational normalised ratio decreased
Blood bilirubin unconjugated increasedIschaemic hepatitis
Blood cholinesterase abnormalJaundice
Blood cholinesterase decreasedJaundice cholestatic
Blood fibrinogen abnormalJaundice hepatocellular
Blood fibrinogen decreasedKayser-Fleischer ring
Blood thrombin abnormalLeucine aminopeptidase increased
Blood thrombin decreasedLiver and small intestine transplant
Blood thromboplastin abnormalLiver disorder
Blood thromboplastin decreasedLiver function test abnormal
Bromosulphthalein test abnormalLiver induration
Caput medusaeLiver injury
Child-Pugh-Turcotte score increasedLiver operation
CholaemiaLiver palpable subcostal
CholestasisLiver sarcoidosis
Cholestatic liver injuryLiver scan abnormal
Cholestatic pruritusLiver tenderness
Chronic hepatic failureLiver transplant
Chronic hepatitisLupoid hepatic cirrhosis
Coagulation factor IX level abnormalLupus hepatitis
Coagulation factor IX level decreasedMitochondrial aspartate aminotransferase increased
Coagulation factor V level abnormalMixed liver injury
Coagulation factor V level decreasedMolar ratio of total branched-chain amino acid to tyrosine
Coagulation factor VII level abnormalNodular regenerative hyperplasia
Coagulation factor VII level decreasedOcular icterus
Coagulation factor X level abnormalOedema because of hepatic disease
Coagulation factor X level decreasedOesophageal varices haemorrhage
Coagulation factor decreasedPerihepatic discomfort
Coma hepaticPneumobilia
Cryptogenic cirrhosisPortal hypertension
Cytolytic hepatitisPortal triaditis
Duodenal varicesPortal vein flow decreased
Foetor hepaticusPortal vein pressure increased
Galactose elimination capacity test abnormalPortopulmonary hypertension
Galactose elimination capacity test decreasedProtein C decreased
Gamma-glutamyltransferase abnormalProtein S abnormal
Gamma-glutamyltransferase increasedProtein S decreased
Gastric varicesProthrombin level abnormal
Granulomatous liver diseaseProthrombin level decreased
Guanase increasedProthrombin time abnormal
Haemorrhagic ascitesProthrombin time prolonged
Hepaplastin abnormalProthrombin time ratio abnormal
Hepaplastin decreasedProthrombin time ratio decreased
HepatectomyRadiation hepatitis
Hepatic atrophyRenal and liver transplant
Hepatic calcificationRetinol binding protein decreased
Hepatic cirrhosisRetrograde portal vein flow
Hepatic congestionReye's syndrome
Hepatic encephalopathySpider naevus
Hepatic encephalopathy prophylaxisSubacute hepatic failure
Hepatic enzyme abnormalThrombin time abnormal
Hepatic enzyme decreasedThrombin time prolonged
Hepatic enzyme increasedTotal bile acids increased
Hepatic failureTransaminases abnormal
Hepatic fibrosisTransaminases increased
Hepatic function abnormalUltrasound liver abnormal
Hepatic hydrothoraxUrine bilirubin increased
Hepatic infiltration eosinophilicUrobilin urine present
Hepatic lesionVarices oesophageal
Hepatic massX-ray hepatobiliary abnormal
Hepatic necrosisYellow skin
Hepatic pain 
Selected skin-related AEs
BlisterIschaemic ulcer
Blister infectedMicroangiopathy
Dermatitis exfoliativeNecrosis ischaemic
Dermatitis exfoliative generalisedNeuropathic ulcer
Diabetic bullosisPain of skin
Diabetic footPeripheral coldness
Diabetic foot infectionPeripheral ischaemia
Diabetic gangrenePeripheral vascular disorder
Diabetic microangiopathyPoor peripheral circulation
Diabetic neuropathic ulcerRaynaud's phenomenon
Diabetic ulcerSkin exfoliation
Diabetic vascular disorderSkin lesion
Dry gangreneSkin necrosis
Exfoliative rashSkin ulcer
Extremity necrosisThrombophlebitis
Infected skin ulcerUlcer
Intermittent claudicationVaricose ulceration
Ischaemic limb painVenous ulcer pain
Selected pancreatitis-related AEs
Blood amylase abnormalPancreatic enzymes increased
Blood amylase increasedPancreatic haemorrhage
Blood trypsin increasedPancreatic necrosis
Cullen's signPancreatic phlegmon
Hereditary pancreatitisPancreatic pseudocyst
HyperlipasaemiaPancreatic pseudocyst drainage
Ischemic pancreatitisPancreatitis
Lipase abnormalPancreatitis acute
Lipase increasedPancreatis haemorrhagic
Oedematous pancreatitisPancreatitis necrotizing
Pancreatic abscessPancreatitis relapsing
Pancreatic enzyme abnormalityPancreatorenal syndrome
Pancreatic enzymes abnormal 
Table A3.  Subject-years of exposure (SYE) in pooled populations.
 Vildagliptin 50 mg qdVildagliptin 50 mg bidAll Comparators
  1. SYE (years) for each treatment is calculated as (duration of exposure in days, disregarding any treatment interruptions, for all subjects in that treatment group)/365.25.

  2. *In studies followed by an extension, for example where patients switched from placebo in the core to vilda 50 qd in the extension, these patients would contribute to both vilda 50 qd and All comparators group in the All studies (excl / incl open-label) safety population.

All studies (including open-label) safety population*
 Patient (N)220661597082
 Exposure (SYE)1385.37351.16705.4
All controlled studies (including open-label) safety population
 Patient (N)150261167082
 Exposure (SYE)737.47034.26705.3
All studies (excluding open-label) safety population*
 Patient (N)204961166210
 Exposure (SYE)1253.97313.66512.7
All controlled studies (excluding open-label) safety population
 Patient (N)150261166210
 Exposure (SYE)737.47034.26512.7