Dipeptidyl peptidase‐4 (DPP‐4) inhibitors for type 2 diabetes mellitus

Bernd Richter; Elizabeth Bandeira‐Echtler; Karla Bergerhoff; Christian Lerch

doi:10.1002/14651858.CD006739.pub2

. 2008 Apr 23;2008(2):CD006739. doi: 10.1002/14651858.CD006739.pub2

Dipeptidyl peptidase‐4 (DPP‐4) inhibitors for type 2 diabetes mellitus

Bernd Richter ^1,^✉, Elizabeth Bandeira‐Echtler ¹, Karla Bergerhoff ¹, Christian Lerch ¹

Editor: Cochrane Metabolic and Endocrine Disorders Group

PMCID: PMC8985075 PMID: 18425967

Abstract

Background

In type 2 diabetes mellitus there is a progressive loss of beta‐cell function. One new approach yielding promising results is the use of the orally active dipeptidyl peptidase‐4 (DPP‐4) inhibitors like sitagliptin and vildagliptin.

Objectives

To assess the effects of dipeptidyl peptidase‐4 (DPP‐4) inhibitors for type 2 diabetes mellitus.

Search methods

Studies were obtained from computerised searches of MEDLINE, EMBASE and The Cochrane Library.

Selection criteria

Studies were included if they were randomised controlled trials in adult people with type 2 diabetes mellitus and had a trial duration of at least 12 weeks.

Data collection and analysis

Two authors independently assessed risk of bias and extracted data. Pooling of studies was performed by means of fixed‐effect meta‐analysis.

Main results

Twenty‐five studies of good quality were identified, 11 trials evaluated sitagliptin and 14 trials vildagliptin treatment. Altogether, 6743 patients were randomised in sitagliptin and 6121 patients in vildagliptin studies, respectively. Sitagliptin and vildagliptin studies ranged from 12 to 52 weeks duration. No data were published on mortality, diabetic complications, costs of treatment and health‐related quality of life.   Sitagliptin and vildagliptin therapy in comparison with placebo resulted in an HbA1c reduction of approximately 0.7% and 0.6%, respectively. Data on comparisons with active comparators were limited but indicated no improved metabolic control following DPP‐4 intervention in contrast to other hypoglycaemic agents. Sitagliptin and vildagliptin therapy did not result in weight gain but weight loss was more pronounced following placebo interventions. No definite conclusions could be drawn from published data on sitagliptin and vildagliptin effects on measurements of beta‐cell function. Overall, sitagliptin and vildagliptin were well tolerated, no severe hypoglycaemia was reported in patients taking sitagliptin or vildagliptin. All‐cause infections increased significantly after sitagliptin treatment but did not reach statistical significance following vildagliptin therapy. All published randomised controlled trials of at least 12 weeks treatment with sitagliptin and vildagliptin only reported routine laboratory safety measurements

Authors' conclusions

DPP‐4 inhibitors have some theoretical advantages over existing therapies with oral antidiabetic compounds but should currently be restricted to individual patients. Long‐term data especially on cardiovascular outcomes and safety are urgently needed before widespread use of these new agents. More information on the benefit‐risk ratio of DPP‐4 inhibitor treatment is necessary especially analysing adverse effects on parameters of immune function. Also, long‐term data are needed investigating patient‐oriented parameters like health‐related quality of life, diabetic complications and all‐cause mortality.

Plain language summary

Dipeptidyl peptidase‐4 (DPP‐4) inhibitors for type 2 diabetes mellitus

Dipeptidyl peptidase‐4 (DPP‐4) inhibitors like sitagliptin and vildagliptin are promising new medicines for the treatment of type 2 diabetes mellitus. They are supposed to improve metabolic control (as measured by lowering blood glucose) without causing severe hypoglycaemia (low blood sugar levels leading to unconsciousness and other symptoms).   Altogether 12.864 people took part in 25 studies investigating the new compounds sitagliptin and vildagliptin. Most studies lasted 24 weeks, the longest trials evaluated 52 weeks of treatment. So far, no study reported on patient‐oriented parameters like mortality, diabetic complications, costs of treatment and health‐related quality of life. When compared to placebo treatment sitagliptin and vildagliptin improved metabolic control. Comparison with other already established blood‐glucose lowering drugs did not reveal advantages of DPP‐4 treatment. Weight gain was not observed after sitagliptin and vildagliptin therapy. Overall, sitagliptin and vildagliptin were well tolerated, no severe hypoglycaemia was reported in patients taking sitagliptin or vildagliptin. However, all‐cause infections increased significantly after sitagliptin treatment but did not reach statistical significance following vildagliptin therapy. Unfortunately, all published randomised controlled trials of at least 12 weeks treatment with sitagliptin and vildagliptin only reported routine laboratory safety measurements. Since the new DPP‐4 inhibitors may influence immune function additional long‐term data on the safety of these drugs are necessary. Also, cardiovascular outcomes like heart attacks and strokes should not be increased with any antidiabetic therapy but data so far are lacking. Until new information arrives, DPP‐4 inhibitors should only be used under controlled conditions and in individual patients.

Summary of findings

Background

Description of the condition

Diabetes mellitus is a metabolic disorder resulting from a defect in insulin secretion, insulin action, or both. A consequence of this is chronic hyperglycaemia (that is elevated levels of plasma glucose) with disturbances of carbohydrate, fat and protein metabolism. Long‐term complications of diabetes mellitus include retinopathy, nephropathy and neuropathy. The risk of cardiovascular disease is increased. For a detailed overview of diabetes mellitus, please see under 'Additional information' in the information on the Metabolic and Endocrine Disorders Group in The Cochrane Library (see 'About', 'Cochrane Review Groups (CRGs)'). For an explanation of methodological terms, see the main glossary in The Cochrane Library.

There are two main types of diabetes mellitus, type 1 (formerly termed insulin‐dependent diabetes mellitus) and type 2 (formerly termed non‐insulin dependent diabetes mellitus):

Type 1 diabetes mellitus

Type 1 diabetes is a chronic disease characterised by hyperglycaemia due to absolute deficiency of insulin secretion which is caused by autoimmune destruction of the pancreatic beta cells. Evidence of autoimmunity is provided by the appearance of autoantibodies prior to the onset of clinical disease. The clinical presentation ranges from mild nonspecific symptoms or no symptoms to coma. Although type 1 diabetes usually develops before 30 years of age, it can occur at any age. At presentation, most patients are thin and have experienced weight loss, polyuria, polydipsia, fatigue, and diabetic ketoacidosis.

Type 2 diabetes mellitus

In type 2 diabetes mellitus, the actions and secretion of insulin are impaired, as opposed to the absolute deficiency of insulin that occurs with type 1 diabetes mellitus. Type 2 diabetes is characterised by two major pathophysiologic defects: (1) insulin resistance, which results in increased hepatic glucose production and decreased peripheral glucose disposal, (2) impaired b‐cell secretory function (Kahn 1997). Insulin resistance is an impaired biological response to the effects of exogenous or endogenous insulin. Insulin resistance in the hepatic and peripheral tissues, particularly skeletal muscle, leads to unrestrained hepatic glucose production and diminished insulin‐stimulated peripheral glucose uptake and utilization (DeFronzo 1992). Insulin secretion by the pancreatic beta cell is initially sufficient to compensate for insulin resistance, thereby maintaining normal blood glucose levels. Hyperinsulinaemia, which accompanies insulin resistance, can maintain sufficiently normal glucose metabolism as long as pancreatic b‐cell function remains normal. However, in patients who may develop type 2 diabetes, insulin secretion eventually fails, leading to hyperglycaemia and clinical diabetes (Warram 1990). Individuals with type 2 diabetes may have few or no classic clinical symptoms (see above) of hyperglycaemia (Ruige 1997). The difficulty in maintaining metabolic control, for example measured by haemoglobin A1c (HbA1c) over time may be related to several behavioral factors (for example difficulties with healthy eating, exercise, medication regimens) but primarily reflects the underlying progressive decline in b‐cell function (UKPDS‐16 1995).   Type 2 diabetes has traditionally been treated in a stepwise manner, starting with lifestyle modifications (Armour 2004; Gimenez‐Perez 2001; Moore 2005), exercise (Thomas 2001) and later on pharmacotherapy with oral agents. Several classes of oral agents are available for clinical use. These mainly include insulin secretagogues, drugs that delay the absorption of carbohydrates from the gastrointestinal tract, and insulin sensitizers. Over time, many patients with type 2 diabetes will require insulin therapy (Burt 2005; Misso 2005; Richter 2005; Roberts 2005; Royle 2003; Siebenhofer 2004).   Insulin secretagogues: Currently, the sulphonylureas used are mainly glibenclamide (glyburide), glipizide, chlorpropamide, tolbutamide, and glimepiride. These drugs stimulate pancreatic b‐cell insulin secretion by binding to a sulphonylurea receptor. The short‐acting non‐sulphonylurea insulin secretagogues are repaglinide and nateglinide (Black 2003). These are newer agents that also stimulate insulin secretion by binding to the sulphonylurea receptor.   Alpha‐glucosidase inhibitors: Acarbose and miglitol are a‐glucosidase inhibitors. These drugs slow the absorption of carbohydrates, reducing especially postprandial elevations in plasma glucose levels. They do not significantly lower fasting plasma glucose levels but cause a modest reduction in HbA1c (Van de Laar 2005).   Insulin sensitizers: Metformin belongs to the biguanides class (Saenz 2005; Salpeter 2003). It might increase insulin sensitivity in the liver by inhibiting hepatic gluconeogenesis and thereby reducing hepatic glucose production. Metformin also seems to increase peripheral insulin sensitivity by enhancing glucose uptake in the muscle. The thiazolidinediones consist of rosiglitazone and pioglitazone. These substances decrease insulin resistance in muscle and adipose tissue by activating the peroxisome proliferator‐activated receptor γ (PPAR‐γ) which increases production of proteins involved in glucose uptake. They also decrease hepatic glucose production by improving hepatic insulin sensitivity.

Description of the intervention

Type 2 diabetes mellitus can be treated by non‐pharmacological (diet, exercise) and pharmacological means. Insulin, as the natural hormone of the body, might be given as animal (mainly pork or beef) insulin (Richter 2005), genetically constructed 'human' insulin or as insulin‐'analogues' with a modified molecular structure compared to human insulin (Roberts 2005; Siebenhofer 2004). Insulin is currently administered by diabetic people in various ways: Subcutaneous injections, insulin pumps (Misso 2005), and by inhalation (Royle 2003). Oral antidiabetic agents are most often used to treat type 2 diabetes mellitus in its initial stages if lifestyle modifications have failed. The thiazolidinediones rosiglitazone and pioglitazone (Richter 2006) offer new oral treatment options and affect many tissues and parts of the body. In order to evaluate their effects not only on metabolic control in type 2 diabetes mellitus but also on patient‐oriented outcomes like cardiovascular disease, longer‐term studies of at least 24 weeks continuous intake will be critically appraised in this review.

It has been demonstrated that, for a given rise in plasma glucose, the increase in plasma insulin is approximately threefold greater when glucose is administered orally compared with intravenously. This enhancement of insulin release is known as the 'incretin' effect. The peptides glucagon‐like peptide‐1 (7‐36) amide (GLP‐1) and glucose‐dependent insulinotropic peptide (GIP) are the most important incretin hormones. Both act as potent insulinotropic hormones, released by oral glucose, and up to two‐thirds of the insulin normally secreted in relation to a meal are thought to be a result of the actions of these hormones. GLP‐1 has been shown to be essential for normal postprandial glucose homeostasis in humans and its secretion throughout the day is highly correlated to the release of insulin. The insulinotropic effect of GLP‐1 is thought to be glucose‐dependent, and this dependence on blood glucose concentration at or above fasting glucose levels means that GLP‐1 should not cause profound hypoglycaemia. Direct effects of GLP‐1 on beta‐cell growth and survival as well as inhibition of beta‐cell apoptosis have been shown in animal models. GLP‐1 suppresses glucagon secretion in a glucose‐dependent manner and therefore is unlikely to impair the glucagon counter‐regulatory response to hypoglycaemia. GLP‐1 delays gastric emptying and secretion, thus reducing postprandial glucose excursions by delaying nutrient delivery to the small intestine (Drucker 2006).   It is recognized that inadequate secretions of insulin are a very early element in the development of type 2 diabetes mellitus and the progression happens because of declining beta‐cell function which in part is a result of loss of beta‐cells. Patients with type 2 diabetes show an almost complete loss of the incretin effect. GLP‐1 is metabolised very rapidly in the circulation, with a half‐life in vivo of less than two minutes. The ubiquitous enzyme, dipeptidyl‐peptidase 4 (DPP‐4) is in many tissues, including the kidney, intestine and capillary endothelium. DPP‐4 metabolises GLP‐1 and is involved in regulating the biological activity of GLP‐1. DPP‐4 inhibitors or 'gliptins' prevent GLP‐1 degradation and improve circulation time, thereby increasing the biological activity of incretin hormones. Two inhibitors are currently on the market, vildagliptin and sitagliptin for once daily oral administration (Drucker 2006).   DPP‐4, also known as the lymphocyte cell surface marker CD 26, is a ubiquitous complex enzyme that exists as a membrane‐spanning cell‐surface aminopeptidase that transmits intracellular signals via a short intracellular tail and a second smaller soluble form circulating in the plasma. It is widely expressed in many tissues, such as liver, lung, kidney, intestine, lymphocytes, capillary endothelium and T‐cells, B‐cells and natural killer cells (Thornberry 2007).

Adverse effects of the intervention

In addition to stabilising the incretins, GLP‐1 and GIP, DPP‐4 inhibitors also prolong the action of a number of neuropeptide like neuropeptide Y, growth hormone‐releasing hormone and chemokinines such as stromal‐cell derived factor 1 and macrophage‐derived chemokine. Potential side‐effects include neurogenic inflammation, increase in blood pressure, enhanced inflammation and allergic reactions. DPP‐4 contributes to T‐cell activation, raising the possibility that these compounds compromise immune function. Therefore, the long‐term safety of DPP‐4 inhibitors merits careful consideration and investigation. The elucidation of several new members of the DPP‐4 family have consequences for the development of DPP‐4 inhibitors. Compound previously though to be specific could in fact be inhibitors of other members of the DPP‐4 enzyme family (Green 2006).   As a T‐cell co stimulator, DPP‐4 is of importance in the immune system. Levels of tissue DPP‐4 are reduced in nasal tissue of people with chronic rhinosinusitis and DPP‐4 inhibition seems to aggravate nasopharyngitis as could be observed in clinical studies. Therefore, it seems to be highly important to monitor DPP‐4 treated patients for the development of inflammatory conditions, such as angioedema, rhinitis and urticaria. DPP‐4 also regulates migration of human cord blood CD34+ progenitor cells and the homing and engraftment of hematopoetic stem cells.

Why it is important to do this review

One systematic review and meta‐analysis on the efficacy and safety of incretin therapy was recently published (Amori 2007). A Cochrane review on glucagon‐like peptide analogues for type 2 diabetes mellitus is currently being accomplished (Snaith 2007). This Cochrane review focuses on DPP‐4 inhibitor compounds and aggregates the most recent data on sitagliptin and vildagliptin therapy. Before widespread use of these new drugs it is necessary to establish an unbiased benefit‐risk ratio to provide guidance to clinicians in the growing market of oral antihypoglycaemic compounds

Objectives

To assess the effects of dipeptidyl peptidase‐4 (DPP‐4) inhibitors for type 2 diabetes mellitus.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled clinical trials.

Types of participants

Adults (over 18 years of age) with type 2 diabetes.   To be consistent with changes in classification and diagnostic criteria of type 2 diabetes mellitus through the years, the diagnosis should have been established using the standard criteria valid at the time of the beginning of the trial (for example ADA 1997; ADA 1999; WHO 1980; WHO 1985; WHO 1998). Ideally, diagnostic criteria should have been described. If necessary, authors' definition of diabetes mellitus will be used. Diagnostic criteria will be eventually subjected to a sensitivity analysis.

Types of interventions

Treatment for a minimum of 12 weeks with DPP‐4 inhibitors (sitagliptin or vildagliptin) alone or in combination with meglitinide analogues, metformin, a sulphonylurea or a thiazolidinedione. The following comparisons will be acceptable for evaluation:

sitagliptin or vildagliptin versus placebo;
sitagliptin or vildagliptin versus single hypoglycaemic agents;
sitagliptin or vildagliptin in combination with other hypoglycaemic agents versus other combinations of hypoglycaemic agents;
sitagliptin or vildagliptin versus intensive lifestyle interventions.

Types of outcome measures

Primary outcomes

metabolic control as measured by glycosylated haemoglobin A1c (HbA1c);
adverse events (for example hypoglycaemia, infections, congestive heart failure, oedema);
health‐related quality of life (using a validated instrument).

Secondary outcomes

weight gain or weight loss (as measured by kg or body mass index (BMI));
beta‐cell function and in particular whether it is preserved over time;
mortality (all‐cause mortality; diabetes related mortality (death from myocardial infarction, stroke, peripheral vascular disease, renal disease, hyper‐ or hypoglycaemia or sudden death));
morbidity (all‐cause morbidity as well as diabetes and cardiovascular related morbidity, for example angina pectoris, myocardial infarction, stroke, peripheral vascular disease, neuropathy, retinopathy, nephropathy, erectile dysfunction, amputation);
costs.

Covariates, effect modifiers and confounders

age;
race;
sex;
compliance;
co‐morbidities (for example myocardial infarction, stroke);
co‐medication (for example antihypertensive drugs, aspirin).

Timing of outcome measurement

Outcomes were assessed in the short (equal or more than 12 weeks to less than 18 weeks), medium term (equal or more than 18 weeks to one year) and long term (more than one year).

Search methods for identification of studies

Electronic searches

We used the following sources for the identification of trials:

The Cochrane Library (issue 1, 2008);
MEDLINE (until Jan 2008);
EMBASE (until Jan 2008).

We also planned to search databases of ongoing trials: Current Controlled Trials (www.controlled‐trials.com ‐ with links to other databases of ongoing trials). This will be performed in future updates of this review.

For detailed search strategy please see under Appendix 1.

Searching other resources

We planned to search the following additional sources:

information on unpublished trials from pharmaceutical companies;
the web sites of the Food and Drug Administration (FDA) and the European Medicines Agency (EMEA);
the web sites of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD).

This will be performed in future updates of this review.

We tried to identify additional studies by searching the reference lists of included trials and (systematic) reviews, meta‐analyses and health technology assessment reports noticed.

Additional key words of relevance could have been detected during any of the electronic or other searches. If this is the case, electronic search strategies would have been modified to incorporate these terms. It was not necessary to add additional key words. Studies published in any language were planned to be included.

Data collection and analysis

Selection of studies

To determine the studies to be assessed further, two authors independently scanned the abstract, titles or both sections of every record retrieved. All potentially relevant articles were investigated as full text. Interrater agreement for study selection was measured using the kappa statistic (Cohen 1960). Differences were planned to be marked and if these studies were later on to be included, the influence of the primary choice would have been subjected to a sensitivity analysis. Where differences in opinion existed, they were resolved by a third party. If resolving disagreement was not possible, the article was to be added to those 'awaiting assessment' and authors would have been contacted for clarification. An adapted PRISMA (Preferred Reporting Items for Systematic Reviews and Meta‐Analyses) flow‐chart of study selection will be attached (Moher 1999).

Data extraction and management

For studies that fulfil inclusion criteria, two authors independently abstracted relevant population and intervention characteristics using standard data extraction templates (for details see Characteristics of included studies and Appendix 2 to Appendix 24) with any disagreements resolved by discussion, or if required by a third party. Any relevant missing information on the trial was planned to be sought from the original author(s) of the article, if required.

Dealing with duplicate publications

In the case of duplicate publications and companion papers of a primary study, we tried to maximise yield of information by simultaneous evaluation of all available data. In cases of doubt, the original publication (usually the oldest version) obtained priority.

Assessment of risk of bias in included studies

Two authors assessed risk of bias of each trial independently. Possible disagreement was resolved by consensus, or with consultation of a third party in case of disagreement. We planned to explore the influence of individual quality criteria in a sensitivity analysis (see under 'sensitivity analyses'). Interrater agreement for key quality indicators (for example concealment of allocation) was calculated using the kappa statistic (Cohen 1960). In cases of disagreement, the rest of the group would have been consulted and a judgement would have been made based on consensus.

Measures of treatment effect

Dichotomous data

Dichotomous outcomes (for example severe hypoglycaemia yes/no) were expressed as odds ratios (OR) or relative risks (RR) with 95% confidence intervals (CI).

Continuous data

Continuous outcomes (for example metabolic control as measured by glycosylated haemoglobin A1c (HbA1c) was expressed as mean differences with 95% CI.

Time‐to‐event data

Time‐to‐event outcomes (for example time until development of diabetic retinopathy) were planned to be expressed as hazard ratios (HR) with 95% CI.

Unit of analysis issues

It was planned to evaluate unit of analysis issues like cluster‐randomised trials, cross‐over trials or multiple observations for the same outcome.

Dealing with missing data

Relevant missing data would have been obtained from authors, if feasible. Evaluation of important numerical data such as screened, eligible and randomised patients as well as intention‐to‐treat (ITT) and per‐protocol (PP) population was carefully performed. Attrition rates, for example drop‐outs, losses to follow‐up and withdrawals were investigated. Issues of missing data, ITT and PP were critically appraised and compared to specification of primary outcome parameters and power calculation.

Assessment of heterogeneity

In the event of substantial clinical or methodological or statistical heterogeneity, study results were not combined by means of meta‐analysis. Heterogeneity was identified by visual inspection of the forest plots, by using a standard χ2‐test and a significance level of α = 0.1, in view of the low power of such tests. Heterogeneity was specifically examined with I² (Higgins 2002), where I² values of 50% and more indicate a substantial level of heterogeneity (Higgins 2003). When heterogeneity was found, we attempted to determine potential reasons for it by examining individual study characteristics and those of subgroups of the main body of evidence.

Assessment of reporting biases

Funnel plots were planned to be used in an exploratory data analysis to assess for the potential existence of small study bias. There are a number of explanations for the asymmetry of a funnel plot, including true heterogeneity of effect with respect to study size, poor methodological design of small studies and publication bias (Sterne 2001). Thus, we carefully used this instrument (Lau 2006).

Data synthesis

Data were summarised statistically if they were available, sufficiently similar and of sufficient quality. Statistical analysis were performed according to the statistical guidelines referenced in the newest version of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2005).

Subgroup analysis and investigation of heterogeneity

Subgroup analyses were planned to be mainly performed if one of the primary outcome parameters demonstrated statistically significant differences between intervention groups. In any other case subgroup analyses were planned to be clearly marked as a hypothesis generating exercise.

The following subgroup analyses were planned:

gender (female versus male);
age (depending on data but especially older versus younger patients);
patients with or without co‐morbidities (for example heart attack, stroke, peripheral vascular disease);
patients with or without co‐medication (for example antihypertensive drugs, statins, aspirin).

Subgroup analyses were planned to be mainly used to explore clinical or methodological or statistical heterogeneity.

Sensitivity analysis

We planned to perform sensitivity analyses in order to explore the influence of the following factors on effect size:

repeating the analysis excluding unpublished studies;
repeating the analysis taking account of study quality, as specified above;
repeating the analysis excluding any very long or large studies to establish how much they dominate the results;
repeating the analysis excluding studies using the following filters: diagnostic criteria, language of publication, source of funding (industry versus other), country.

The robustness of the results was also planned to be tested by repeating the analysis using different measures of effects size (relative risk, odds ratio etc.) and different statistical models (fixed and random‐effects models).

Results

Description of studies

Results of the search

The initial search identified 1083 records, from these, 52 full papers were identified for further examination. The other studies were excluded on the basis of their abstracts because they were not relevant to the question under study or clearly did not meet inclusion/exclusion criteria (see Figure 1 for details of the amended QUOROM (quality of reporting of meta‐analyses) flow chart for study selection). After screening the full text of the selected papers, 25 studies published in 27 publications finally met the inclusion criteria, 25 publications were (systematic) reviews or overviews (Ahren 2003; Ahren 2006; Ahren 2007; Amori 2007; Barnett 2006; Campbell 2007; Canadian 2006; Deacon 2005; Drucker 2006; Drucker 2007; Gallwitz 2007; Henness 2006; Herman 2007; Idris 2007; Kleppinger 2007; Levetan 2007; Lyseng‐William. 2007; Mest 2006; Miller 2006; Pratley 2007; Ristic 2006; Schlesselman 2006; Sebokova 2007; Thornberry 2007; Todd 2007). All studies were published in English.

Amended QUOROM (quality of reporting of meta‐analyses) flow chart of study selection

Of the 25 included trials, 11 studies analysed sitagliptin treatment (Aschner 2006; Charbonnel 2006; Goldstein 2007; Hanefeld 2007; Hermansen 2007; Nauck 2007; Nonaka 2008; Raz 2006; Rosenstock 2006; Scott 2007a; Scott 2007b), 14 studies evaluated vildagliptin therapy (Ahren 2004; Bolli 2008; Bosi 2007; Dejager 2007; Fonseca 2007; Garber 2007; Mimori 2006; Pi‐Sunyer 2007; Pratley 2006; Ristic 2005; Rosenstock 2007a; Rosenstock 2007b; Scherbaum 2008; Schweizer 2007).

Interrater agreement

Interrater agreement for study selection, that is qualifying a study as 'included' or 'potentially' relevant was complete with no reference necessary to be discussed by a third author.

Included studies

For full details please note the table Characteristics of included studies

Interventions and comparisons

Sitagliptin

Comparisons with placebo

Six trials or study arms compared sitagliptin monotherapy with placebo:

Aschner 2006: sitagliptin 100 or 200 mg daily versus placebo;
Goldstein 2007: sitagliptin 100 mg daily versus placebo;
Hanefeld 2007: sitagliptin 25 mg or 50 mg or 100 mg daily versus placebo;
Nonaka 2008: sitagliptin 100 mg daily versus placebo;
Raz 2006: sitagliptin 100 mg or 200 mg daily versus placebo;
Scott 2007a: sitagliptin 10 mg or 25 mg or 50 mg or 100 mg daily versus placebo.

Comparisons with single hypoglycaemic agents

Two trials or study arms compared sitagliptin monotherapy with another hypoglycaemic agent monotherapy:

Goldstein 2007: sitagliptin 100 daily mg versus metformin 1000 mg or 2000 mg daily;
Scott 2007a: sitagliptin 100 mg daily versus glipizide 5mg to 20 mg daily.

Comparisons of combination therapies

Six trials or study arms compared sitagliptin combination therapies with other combination therapies of hypoglycaemic agents:

Charbonnel 2006: sitagliptin 100 mg daily (add‐on to metformin therapy) versus placebo (add‐on to metformin therapy);
Goldstein 2007: sitagliptin 50 mg daily plus metformin 1000 or 2000 mg daily versus metformin 1000 or 2000 mg daily;
Hermansen 2007: sitagliptin 100 mg daily (add‐on to ongoing stable doses of glimepiride, alone or in combination with metformin) versus placebo (add‐on to ongoing stable doses of glimepiride, alone or in combination with metformin);
Nauck 2007: sitagliptin 100 mg daily (add‐on to metformin therapy) versus placebo (add‐on to metformin therapy);
Rosenstock 2006: sitagliptin 100 mg daily (add‐on to pioglitazone therapy) versus placebo (add‐on to pioglitazone therapy);
Scott 2007b: sitagliptin 100 mg daily (add‐on to metformin therapy) versus rosiglitazone 8 mg (add‐on to metformin therapy) versus placebo (add‐on to metformin therapy).

Vildagliptin

Comparisons with placebo

Six trials or study arms compared vildagliptin monotherapy with placebo:

Dejager 2007: vildagliptin 50 mg or 100 mg daily versus placebo;
Mimori 2006: vildagliptin 20 mg or 50 mg or 100 mg daily versus placebo;
Pi‐Sunyer 2007: vildagliptin 50 mg or 100 mg daily versus placebo;
Pratley 2006: vildagliptin 50 mg daily versus placebo;
Ristic 2005: vildagliptin 25 mg or 50 mg or 100 mg daily versus placebo;
Scherbaum 2008: vildagliptin 50 mg daily versus placebo.

Comparisons with single hypoglycaemic agents

Three trials or study arms compared vildagliptin monotherapy with another hypoglycaemic agent monotherapy:

Rosenstock 2007a: vildagliptin 100 mg daily versus rosiglitazone 8 mg daily;
Rosenstock 2007b: vildagliptin 100 mg daily versus pioglitazone 30 mg daily;
Schweizer 2007: vildagliptin 100 mg daily versus metformin up to 2000 mg daily.

Comparisons of combination therapies

Six trials or study arms compared vildagliptin combination therapies with other combination therapies of hypoglycaemic agents:

Ahren 2004: vildagliptin 50 mg daily (add‐on to metformin therapy) versus placebo (add‐on to metformin therapy)
Bolli 2008: vildagliptin 100 mg daily (add‐on to metformin therapy) versus pioglitazone 30 mg daily (add‐on to metformin therapy)
Bosi 2007: vildagliptin 50 or 100 mg daily (add‐on to metformin therapy) versus placebo (add‐on to metformin therapy)
Fonseca 2007: vildagliptin 100 mg daily (add‐on to insulin therapy) versus placebo (add‐on to insulin therapy)
Garber 2007: vildagliptin 50 or 100 mg daily (add‐on to pioglitazone therapy) versus placebo (add‐on to pioglitazone therapy)
Rosenstock 2007b: vildagliptin 50 mg or 100 mg daily plus 15 mg or 30 mg pioglitazone daily versus pioglitazone 30 mg daily

Number of study centres, countries, location and setting

With the exception of two Japanese studies (Mimori 2006; Nonaka 2008) all trials had a multinational design. Only one publication described the number of study centres with regard to sitagliptin trials (Nonaka 2008). Almost all publications of vildagliptin therapy specified number of study centres ranging from 15 to 202. In sitagliptin and vildagliptin studies 1 to 34 and 1 to 11 countries participated. Sitagliptin was mainly investigated in Asia, Europe, North and South America whereas vildagliptin trials mainly took place in Europe and North America. Only one study described the setting (outpatients) in which the trial was performed (Ristic 2005).

Treatment before study

Most sitagliptin studies allowed pre‐treatment with oral antihyperglycaemic drugs, whereas approximately half of the vildagliptin trials started with drug‐naive patients, that is patients who were treated only by diet, exercise or both.

Duration of the intervention, run‐in period

Sitagliptin studies ranged from 12 to 52 weeks duration, most trials lasted 24 weeks with a single trial (Nauck 2007) of 52 weeks duration. Vildagliptin studies showed a comparable distribution with two trials lasting 52 weeks (Scherbaum 2008; Schweizer 2007). Altogether 6743 patients were randomised in sitagliptin trials and 6121 in vildagliptin trials, respectively.

Participants, inclusion and exclusion criteria, diagnostic criteria, co‐morbidities and medications

In sitagliptin trials most patients were inadequately controlled (rarely defined), either on diet, exercise or both or on metformin (Charbonnel 2006; Nauck 2007; Scott 2007b), glimepiride with or without metformin (Hermansen 2007) or pioglitazone (Rosenstock 2006) treatment.   Most vildagliptin trials investigated drug‐naive patients (Dejager 2007; Mimori 2006; Pi‐Sunyer 2007; Pratley 2006; Ristic 2005; Rosenstock 2007a; Rosenstock 2007b; Scherbaum 2008; Schweizer 2007). The other studies examined patients with inadequate control (rarely defined) on metformin (Ahren 2006; Bolli 2008; Bosi 2007), insulin (Fonseca 2007) or thiazolidinedione (Garber 2007) treatment   Diagnostic criteria as an inclusion standard were rarely defined, most exclusion criteria consisted of significant diseases such as cardiovascular, liver or renal disorders. No publication offered relevant information about co‐morbidities or co‐medications.

Primary outcomes

For details on outcome data see Appendix 2 and Appendix 3.     All sitagliptin and vildagliptin studies defined glycosylated haemoglobin A1c or change in HbA1c from baseline to endpoint as the primary outcome (with the exception of Mimori 2006 which was only published as an abstract).

Secondary outcomes, additional/other outcomes

Secondary outcomes in sitagliptin and vildagliptin studies consisted mainly of fasting plasma glucose and lipids, insulin, proinsulin, measurements of beta‐cell function, such as proinsulin‐to‐insulin ratio, homeostasis model assessment beta (HOMA‐beta) and insulinogenic index at peak glucose (I/G) as well as the corrected insulin response at peak glucose (CIR(GluPeak), measurements of insulin resistance, such as HOMA‐insulin resistance (IR) and the quantitative insulin sensitivity check index (QUICKI). Moreover, standard meal tolerance tests were performed, for example to analyse plasma glucose, insulin, C‐peptid, 2‐hr post‐prandial glucose, area‐under‐the‐curve (AUC) insulin, C‐peptide AUC and insulin AUC‐to‐glucose AUC ratio. Some studies also evaluated responders to treatment, defined as the percentage of patients achieving certain HbA1c endpoints A1C (for example, less than 7%).   Safety outcomes mainly comprised adverse experiences including hypoglycaemic episodes, physical examinations, electrocardiograms (ECG) and body weight. Laboratory measurements were composed of routine or standard haematology, serum chemistry and urinalysis.

Excluded studies

No study (apart from 25 (systematic) reviews or overviews) had to be excluded after careful evaluation of the full publication.

Risk of bias in included studies

For details on study populations like numbers randomised, analysed, intention‐to‐treat and safety population see Table 3, Figure 2 and Figure 3.

1. Study populations.

Study ID	[n] randomised	[n] safety	[n] ITT	[n] finished study	comments
SITAGLIPTIN STUDIES
Aschner 2006	741	741	711	639	efficacy analyses were based on the all‐patients ‐treated population
Charbonnel 2006	701	701	677	608	efficacy analyses were based on the all‐patients‐treated population;   safety analyses were performed using the all‐patients‐as‐treated population (APaT)
Goldstein 2007	1091	1091	1056	906	efficacy analyses were based on the all‐patients‐treated population
Hanefeld 2007	555	552	535	472	efficacy analyses were based on all‐patients‐treated population
Hermansen 2007	441	441	425	364	efficacy analyses were based on the all‐patients treated population;   safety and tolerability analyses were performed in the all‐patients‐as treated population:   all randomized patients were included in the APaT population
Nauck 2007	1172	1172	793	798
Nonaka 2008	152	151	150	140	primary efficacy analysis was based on the all‐patients‐treated population
Raz 2006	521	521	495	463	efficacy analyses were based on the all‐patients‐treated population
Rosenstock 2006	353	353	337	307	efficacy analyses were performed on the all‐patients‐treated population;
Scott 2007a	743	740	725	651	efficacy analyses were based on the all‐patients‐treated population
Scott 2007b	273	272	266	254	efficacy analyses were based on the all‐patients‐treated population

VILDAGLIPTIN STUDIES
Ahren 2004	107	107	107	97
Bolli 2008	576	575	510	506
Bosi 2007	544	541	416	462	intention to treat (ITT) = primary ITT
Dejager 2007	632	625	380	511	ITT ( = primary ITT)
Fonseca 2007	296	296	290	238
Garber 2007	463	462	398	376	ITT ( = primary ITT)
Mimori 2006	291
Pi‐Sunyer 2007	354	352	340	273
Pratley 2006	100	98	98	91
Ristic 2005	279	276	272	nr
Rosenstock 2007a	786	782	697	678	ITT ( = primary ITT)
Rosenstock 2007b	607	606	592	513
Scherbaum 2008	306	nr	302	264
Schweizer 2007	780	771	760	569

Symbols & abbreviations   nr = not reported

Sitagliptin for type 2 diabetes mellitus
Patient or population: type 2 diabetes mellitus Settings: Intervention: Sitagliptin
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect   (95% CI)	No of Participants   (studies)	Quality of the evidence   (GRADE)	Comments
Outcomes	Assumed risk	Corresponding risk	Relative effect   (95% CI)	No of Participants   (studies)	Quality of the evidence   (GRADE)	Comments
	Control	Sitagliptin
Morbidity	See comment	See comment	Not estimable	‐	See comment	Not investigated
Health‐related quality of life	See comment	See comment	Not estimable	‐	See comment	Not investigated
Change in HbA1c from baseline to endpoint   %   (follow‐up: 12 to 52 weeks)	The mean change in hba1c from baseline to endpoint ranged across control groups from   ‐1.31 to 0.41	The mean Change in HbA1c from baseline to endpoint in the intervention groups was   0.54 lower   (0.58 to 0.5 lower)		6907   (11)	⊕⊕⊕⊕   high¹
Change in HbA1c from baseline to endpoint ‐ Sitagliptin versus another single hypoglycaemic agent   %   (follow‐up: 12 to 24 weeks)	The mean change in hba1c from baseline to endpoint ‐ sitagliptin versus another single hypoglycaemic agent ranged across control groups from   ‐1.13 to ‐0.76	The mean Change in HbA1c from baseline to endpoint ‐ Sitagliptin versus another single hypoglycaemic agent in the intervention groups was   0.33 higher   (0.18 to 0.48 higher)		592   (2)	⊕⊕⊕⊕   high¹
Change in HbA1c from baseline to endpoint [%] ‐ Sitagliptin versus placebo   %   (follow‐up: 18 to 52)	The mean change in hba1c from baseline to endpoint [%] ‐ sitagliptin versus placebo ranged across control groups from   0.12 to 0.18	The mean Change in HbA1c from baseline to endpoint [%] ‐ Sitagliptin versus placebo in the intervention groups was   0.75 lower   (0.86 to 0.63 lower)		1103   (3)	⊕⊕⊕⊕   high¹
Change in body weight from baseline to endpoint ‐ Sitagliptin versus placebo   kg   (follow‐up: 18 to 24 weeks)	The mean change in body weight from baseline to endpoint ‐ sitagliptin versus placebo ranged across control groups from   ‐1.1 to ‐0.7	The mean Change in body weight from baseline to endpoint ‐ Sitagliptin versus placebo in the intervention groups was   0.69 higher   (0.32 to 1.06 higher)		1109   (3)	⊕⊕⊕⊕   high
Adverse events ‐ All‐cause infections   (follow‐up: 12 to 52 weeks)	Medium risk population		RR 1.29   (1.09 to 1.52)	3589   (8)	⊕⊕⊕⊕   high
	77 per 1000	99 per 1000   (84 to 117)	RR 1.29   (1.09 to 1.52)	3589   (8)	⊕⊕⊕⊕   high
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).     CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidance   High quality: Further research is very unlikely to change our confidence in the estimate of effect.   Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.   Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.   Very low quality: We are very uncertain about the estimate.

Vildagliptin for type 2 diabetes mellitus
Patient or population: patients with type 2 diabetes mellitus Settings: Intervention: Vildagliptin
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect   (95% CI)	No of Participants   (studies)	Quality of the evidence   (GRADE)	Comments
Outcomes	Assumed risk	Corresponding risk	Relative effect   (95% CI)	No of Participants   (studies)	Quality of the evidence   (GRADE)	Comments
	Control	Vildagliptin
Morbidity	See comment	See comment	Not estimable	‐	See comment	Not investigated in the included studies
Health‐related quality of life	See comment	See comment	Not estimable	‐	See comment	Not investigated in the included studies
Change in HbA1c from baseline to endpoint   %   (follow‐up: 12 to 52 weeks)	The mean change in hba1c from baseline to endpoint ranged across control groups from   ‐1.4 to 0.28 %	The mean Change in HbA1c from baseline to endpoint in the intervention groups was   0.27 lower   (0.29 to 0.26 lower)		6308   (14)	⊕⊕⊕⊕   high¹
Change in HbA1c from baseline to endpoint ‐ Vildagliptin versus another single hypoglycaemic agent   %   (follow‐up: 12 to 52 weeks)	The mean change in hba1c from baseline to endpoint ‐ vildagliptin versus another single hypoglycaemic agent ranged across control groups from   ‐1.4 to ‐1.3	The mean Change in HbA1c from baseline to endpoint ‐ Vildagliptin versus another single hypoglycaemic agent in the intervention groups was   0.3 higher   (0.14 to 0.46 higher)		1764   (3)	⊕⊕⊕⊕   high¹
Change in HbA1c from baseline to endpoint ‐ Vildagliptin versus placebo   %   (follow‐up: 12 to 52 weeks)	The mean change in hba1c from baseline to endpoint ‐ vildagliptin versus placebo ranged across control groups from   ‐0.3 to 0.28	The mean Change in HbA1c from baseline to endpoint ‐ Vildagliptin versus placebo in the intervention groups was   0.32 lower   (0.34 to 0.3 lower)		1139   (6)	⊕⊕⊕⊕   high¹
Change in body weight from baseline to endpoint ‐ Vildagliptin versus placebo   kg   (follow‐up: 12 to 24 weeks)	The mean change in body weight from baseline to endpoint ‐ vildagliptin versus placebo ranged across control groups from   ‐1.4 to ‐0.73	The mean Change in body weight from baseline to endpoint ‐ Vildagliptin versus placebo in the intervention groups was   0.76 higher   (0.19 to 1.32 higher)		484   (3)	⊕⊕⊕⊕   high
Adverse events ‐ All‐cause infections   (follow‐up: 12 to 52 weeks)	Medium risk population		RR 1.04   (0.87 to 1.24)	3573   (10)	⊕⊕⊕⊕   high
	143 per 1000	149 per 1000   (124 to 177)	RR 1.04   (0.87 to 1.24)	3573   (10)	⊕⊕⊕⊕   high
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).     CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidance   High quality: Further research is very unlikely to change our confidence in the estimate of effect.   Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.   Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.   Very low quality: We are very uncertain about the estimate.

Study	Primary outcomes	Secondary outcomes	Additional outcomes	Safety measurements	Laboratory outcomes
Aschner 2006	HbA1c	nr	FPG, insulin, proinsulin, fasting lipids, beta‐cell function: proinsulin‐to‐insulin ratio, HOMA‐beta, insulin resistance: HOMA‐IR, QUICKI, standard meal tolerance test: plasma glucose, insulin, C‐peptid, 2‐hr PPG, AUC insulin AUC, C‐peptide AUC, insulin AUC‐to‐glucose AUC ratio	adverse experiences (prespecified: hypoglycaemia, GI: abdominal pain,nausea, vomiting, diarrhea), physical exams, ECG, body weight	complete blood chemistry,haematology, urinalysis
Charbonnel 2006	change from baseline at week 24 in HbA1c	change from baseline at week 24 in: FPG, glucose, insulin, C‐peptide (after a standard meal) and lipid panel (total cholesterol, triglycerides, LDL‐, HDL‐, non‐HDL cholesterol, triglyceride‐to‐HDL cholesterol ratio	(exploratory endpoints): mean glucose, insulin, C‐peptide, AUC for glucose, insulin, C‐peptide, insulin AUC‐to‐glucose AUC ratio (after standard morning meal)	adverse experiences (special interest: hypoglycaemia, GI AEs), physical exams, vital signs, body weight, ECG	safety lab: routine haematology, serum chemistry, urinalysis
Goldstein 2007	change from baseline at week 24 in HbA1c	nr	change from baseline at week 24in: FPG, fasting serum insuline, fasting serum proinsuline, fasting lipids, beta‐cell function: proinsulin/insulin ratio, HOMA‐beta, insulin resistance: HOMA‐IR, QUICKI (all after standard meal tolerance test)	adverse experiences (prespecified: hypoglycaemia, GI: abdominal pain,nausea, vomiting, diarrhea), physical exams, vital signs, ECG, body weight	complete blood chemistry,haematology, urinalysis
Hanefeld 2007	change from baseline at week 12 in HbA1c		FPG, serum insulin, plasma lipid parameters (total chol., LDL, HDL, triglycerides, FFA), beta‐cell function: HOMA‐beta, insulin resistance: HOMA‐IR, QUICKI, 7‐point home‐glucose measurements = mean daily glucose, % achieving HbA1c <= 7%)	adverse experiences, physical exams, vital signs, ECGs, body weight, hypoglycaemia	routine haematology, serum chemistry, urinalysis
Hermansen 2007	change in HbA1c from baseline to week 24	FPG, 2‐h post‐meal glucose, lipid measurements	meal tolerance test; HOMA‐B, 2‐h postprandial glucose	adverse experiences
Nauck 2007	change from baseline at week 52 in HbA1c (non‐inferiority, per protocol)	nr	FPG, insulin, proinsulin, lipid parameters (total cholesterol, triglycerides, LDL‐, HDL‐, non‐HDL), beta‐cell function: proinsulin/insulin ratio, HOMA‐beta, insulin resistance: HOMA‐IR, QUICKI), durability of treatment: comparing the rate of rise in HbA1c from week 24 to week 52	adverse experiences (prespecified: hypoglycaemia, abdominal pain, nausea, vomiting, diarrhoea), physical exams, vital signs, ECGs, body weight, hypoglycaemia (log book)	blood chemistry, haematology, urinalysis
Nonaka 2008	change in HbA1c from baseline (randomization) at week 12	nr	change from baseline in FPG, 2‐h postprandial glucose (PPG), 1,5‐anhydroglucitol, fasting insulin, fasting serum C‐peptide,   HOMA‐IR, HOMA‐b‐cell function; proportion of patients achieving an HbA1c of <7% or 6.5%:   meal tolerance test in a subgroup of patients	adverse experience reports, vital signs, body weight, ECG, hypoglycaemia (diaries)	hematology, chemistry, urinalysis
Raz 2006	HbA1c	(key secondary endpoints) FPG, insulin, proinsulin, lipids; subset of patients: meal tolerance test ‐> key postprandial endpoints: 2‐h post‐meal glucose, insulin, C‐peptide, 3‐h post‐meal glucose, insulin, C‐peptide and insulin and glucose AUCs		adverse experiences (prespecified: change from baseline in body weight, abdominal pain, nausea, vomiting, diarrhoea), physical exams, vital signs, ECGs, body weight	blood chemistry (includ. ALAT, ASAT, total bilirubin. AP, CK and creatinine), urinalysis
Rosenstock 2006	change from baseline at week 24 in HbA1c	change from baseline in FPG, insulin, proinsulin; beta‐cell function: proinsulin/insulin ratio, HOMA‐beta; insulin resistance: HOMA‐IR, QUICKI; %changes from baseline in selected lipid parameters: total cholesterol, LDL, triglycerides, HDL, non‐HDL‐C; % with HbA1c <7%; proportion requiring rescue therapy		physical exams, vital signs, ECGs, adverse experiences includ. hypoglycaemia and selected GI‐related AEs (abdominal pain, nausea, vomiting, diarrhea)	haematology, serum chemistry (includ. ALAT, AST, total bilirubin, AP), urinalysis
Scott 2007a	HbA1c	nr	change or %change from baseline at week 12: FPG, MDG, MTT‐related variables including 2‐h PPG and glucose AUC, lipid parameters, HOMA‐beta, HOMA‐IR and QUICKI	adverse experiences (AEs of special interest included hypglycaemia (daily glucose logs) and GI‐related symptoms), physical exams, vital signs, ECGs, change in body weight	blood chemistry, haematology, urinalysis
Scott 2007b	HbA1c	nr	FPG, fasting serum   insulin, fasting serum proinsulin, fasting plasma lipids;   beta‐cell function: proinsulin/insulin ratio and HOMA‐b insulin resistance: HOMA of insulin resistance   (HOMA‐IR);   standard meal tolerance test: 2‐h insulin and C‐peptide levels, glycaemic excursion from the 0‐h time point to the 2‐h time point of the MTT (i.e. incremental 2‐h PPG)	adverse experiences, physical examinations, vital signs, body weight	blood chemistry, haematology, urinalysis
Symbols & abbreviations:     nr = not reported   ALT = alanine aminotransferase; AP = alkaline phosphatase; AST = aspartate aminotransferase; AUC = area under the cureve; BMI = body mass index (kg/m2); BP = blood pressure; CK creatine phosphokinase; CRP = C‐reactive protein ECG = electrocardiogram; FP(B)G = fasting plasma (blood) glucose; GI = gastrointestinal; HbA1c = glycosylated haemoglobin A1c; HOMA = homeostasis model assessment (of insulin sensitvity ‐ IR) MDG = mean daily glucose: MTT = meal tolerance test; PPG = postprandial glucose; QUICKI = quantitative insulin sensitivity check index

Characteristic	Aschner 2006	Charbonnel 2006	Goldstein 2007	Hanefeld 2007	Hermansen 2007	Nauck 2007	Nonaka 2008
Intervention 1 (I1) / intervention 2 (I2) / control 1 (C1)	I1:sitagliptin 100mg o.d.   I2: sitagliptin 200mg o.d.   C1: placebo	I1:sitagliptin 100mg o.d.+ metformin >= 1500mg/day   C1: placebo+ metf >= 1500mg/day	I1: sitagliptin 50 mg b.i.d. + metformin 500mg b.i.d.   I2: sitagliptin 50 mg b.i.d.+ metformin 1000md b.i.d.   I3: sitagliptin 100mg o.d.   C1: placebo   C2: metformin 500mg b.i.d.   C3: metformin 1000mg b.i.d.	I1: sitagliptin 25mg o.d.   I2: sitagliptin 50mg o.d.   I3: sitagliptin 100mg o.d.   I4: sitagliptin 50mg b.i.d.   C1: placebo	I1: sitagliptin 100mg o.d.+ glimepiride >= 4mg/day   I2: sitagliptin 100mg o.d. + glimepiride >= 4 mg/day + metformin >= 1500mg/day   C1: placebo + glimepiride >= 4mg/day   C2: placebo + glimepiride >=4mg/day + metformin >= 1500mg/day	I1: sitagliptin 100mg o.d.+ metformin >= 1500 mg/day   C1: glipizide uptitr. 5 ‐ 20mg/day + metformin >= 1500 mg/day	I1: sitagliptin 100mg o.d.   C1: placebo
[n] (I1/ I2 / C1 / total)	I1: 238   I2: 250   C1: 253   Total: 741	I1: 464   C1: 237   Total: 701	I1: 190   I2: 182   I3: 179   C1: 176   C2: 182   C3: 182   Total: 1091	I1: 111   I2: 112   I3: 110   I4: 111   C1: 111   Total: 555	I1: 106   I2: 116   C1: 106   C2: 113   Total: 441	I1: 588   C1: 584   Total: 1172	I1:75   C1: 76   Total: 151
Sex [n,%]	I1: female 102 (42.9); male 136 (57.1)   I2: female 133 (53.2); male 117 (46.8)   C1: female 123 (46.8); male 130 (51.4)	I1: female 205 (44.2); male 259 (55.8)   C1: female 96 (40.5); male 141 (59.5)	I1: female 85 (44.7) ; male 105 (55.3)   I2: female 105 (57.7); male 77 (42.3)   I3: female 86 (48.0); male 93 (52.0)   C1: female 83 (47.2); male 93 (52.8)   C2: female 93 (51.1); male 89 (48.9)   C3: female 100 (54.9); male 82 (45.1)	I1: female 54 (48.6); male 57 (51.4)   I2: female 61 (54.5); male 51 (45.5)   I3: female 49 (44.5); male 61 (55.5)   I4: female 62 (55.9); male 49 (44.1)   C1: female 41 (36.9); male 70 (63.1)	I1: female 50 (47.2); male 56 (52.8)   I2: female 55 (47.4); male 61 (52.6)   C1: female 48 (45.3); male 58 (54.7)   C2: female 54 (47.8); male 59 (52.2)	I1: female 252 (42.9); male 336 (57.1)   C1: female 226 (38.7); male 358 (61.3)	I1: female 30 (40); male 45 (60)   C1: female 26 (34); male 50 (66)
Age [years] mean (SD)	I1: 53.4 (9.5)   I2: 54.9 (10.1)   C1: 54.3 (10.1)	I1: 54.4 (10.4)   C1: 54.7 (9.7)	I1: 54.1 (10.0)   I2: 53.3 (9.6)   I3: 53.3 (10.2)   C1: 53.6 (10.0)   C2: 53.4 (10.2)   C3: 53.2 (9.6)	I1: 55.1 (9.6)   I2: 55.3 (10.3)   I3: 56.0 (7.9)   I4: 55.2 (9.5 )   C1: 55.9 (9.3)	I1: 54.4 (10.3)   I2: 56.5 (8.8)   C1: 55.2 (10.2)   C2: 57.7 (8.9)	I1: 56.8 (9.3)   C1: 56.6 (9.8)	I1: 55.6 (8.6)   C1: 55.0 (8.0)
Ethnic groups [%]	I1: asian 13.4; black 4.2; hispanic 24.4; caucasian: 51.3; other 6.7   I2: asian 14.8: black 4.8; hispanic 21.2; caucasian 52.8; other 6.4   C1: asian 13.4; black 6.3; hispanic 25.3; caucasian 50.2; other 4.7	I1: asian 10.6;   black 6.7; hispanic 15.5; white 63.1; other 4.1   C1: asian 11.0; black 5.9; hispanic 11.8; white 67.1; other 4.2	I1: white 53.7; black 6.8; hispanic 28.9; asian 4.7; other 5.8   I2: white 52.2; black 7.7; hispanic 26.9; asian 6.0; other 7.1   I3: white 52.0; black 6.1; hispanic 29.1; asian 3.4; other 9.5   C1: white 46.0; black 9.7; hispanic 26.7; asian 6.8; other 10.8   C2: white 47.8; black 6.6; hispanic 30.2; asian 7.7; other 7.7   C3: white 58.2; black 4.9; hispanic 21.4; asian 5.5; other 9.9	I1: asian 0.9; black 3.6; white 88.3; other 7.2   I2: asian 0; black 8.0; white 85.7; other 6.3   I3: asian 0; black 5.5; white 88.2; other 6.4   I4: asian 0.9; black 6.3; white 81.1; other 11.7   C1: asian 0.9; black 7.2; white 78.4; other 13.5	I1: caucasian 57.5; black 6.6; hispanic 24.5; asian 5.7; other 5.7   I2: caucasian 64.7; black 2.6; hispanic 11.2; asian 13.8; other 7.8   C1: caucasian 55.7; black 2.8; hispanic 23.6; asian11.3; other 6.6   C2: caucasian 71.7; black 8.0; hispanic 6.2; asian 11.5; other 2.7	I1: caucasian: 73.5; black: 7.0; hispanic 7.3; asian: 8.5; other: 3.7   C1: caucasian: 74.3; black: 6.0; hispanic: 7.9; asian: 8.4; other: 3.4	I1: japanese 100%   C1: japanese 100 %
Duration of disease [years] mean (SD)	I1: 4.3 (4.9)   I2: 4.3 (4.7)   C1: 4.6 (4.7)	I1: 6.0 (5.0)   C1: 6.6 (5.5)	I1: 4.5 (4.7)   I2: 4.4 (4.2)   I3: 4.4 (4.6)   C1: 4.6 (4.9)   C2: 4.5 (3.9)   C3: 4.4 (4.4)	I1: 3.6 (3.4)   I2: 3.3 (3.9)   I3: 3.6 (3.9)   I4: 4.5 (5.9)   C1: 3.3 (3.4)	I1: 7.2 (5.0)   I2: 9.3 (5.7)   C1: 8.0 (6.5)   C2: 10.6 (6.8)	I1: 6.5 (6.1)   C1: 6.2 (5.4)	I1: 4.0 (4.1)   C1: 4.1 (4.6)
Body mass index [kg/m2] mean (SD)	I1: 30.3 (5.2)   I2: 30.3 (5.4)   C1: 30.8 (5.5)	I1: 30.9 (5.3)   C1: 31.5 (4.9)	I1: 32.1 (6.7)   I2: 32.4 (6.6)   I3: 31.2 (5.9)   C1: 32.5 (6.7)   C2: 32.1 (6.8)   C3: 32.2 (7.1)	I1: 31.9 (4.8)   I2: 31.6 (4.9)   I3: 31.6 (5.8)   I4: 32.7 (4.8)   C1: 31.4 (5.1)	I1: 31 (6.7)   I2: 31.3 (5.9)   C1: 30.7 (6.4)   C2: 30.7 (6.2)	I1: 31.2 (5.0)   C1: 31.3 (5.2)	I1: 25.2 (3.5)   C1: 25.1 (3.2)
Pharmaco‐naive patients [n,%]	I1: 124 (52.1)   I2: 125 (50)   C1: 129 (51.0)	I1: 27 (5.8)   C1: 14 (5.9)	I1: 102 (53.7)   I2: 88 ( 48.4)   I3: 91 (50.8)   C1: 88 (50.0)   C2: 91 (50.0)   C3: 90 (49.5)	I1: 41 (36.9)   I2: 39 (34.8)   I3: 41 (37.3)   I4: 38 (34.2)   C1: 39 (35.1)	I1: 11 (10.4)   I2: 0   C1: 8 (7.5)   C2: 3 (2.7)	I1: 25 (4.3)   C1: 28 (4.8)	I1: 36 (48.0)   C1: 29 (38.2)
HbA1c [%] mean (SD)	I1: 8.01 (0.88)   I2: 8.08 (0.94)   C1: 8.03 (0.82)	I1: 7.96 (0.81)   C1: 8.03 (0.82)	I1: 8.8 (1.0)   I2: 8.7 (0.9)   I3: 8.9 (1.0)   C1: 8.7 (1.0)   C2: 8.9 (1.0)   C3: 8.7 (0.9)	I1: 7.7 (0.9)   I2: 7.6 (1.0)   I3: 7.8 (0.9)   I4: 7.8 (0.9)   C1: 7.6 (0.9)	I1: 8.42 (0.79)   I2: 8.27 (0.73)   C1: 8.43 (0.80)   C2: 8.26 (0.68)	I1: 7.7 (0.9)   C1: 7.6 (0.9)	I1: 7.5 (0.9)   C1: 7.7 (0.9)
Notes
Symbols & abbreviations:   Y = yes; N = no; ? = unclear   I = intervention; C = control

Characteristic	Raz 2006	Rosenstock 2006	Scott 2007a	Scott 2007b
Intervention 1 (I1) / intervention 2 (I2) / control 1 (C1)	I1: sitagliptin 100 mg o.d.   I2: sitagliptin 200mg o.d.   C1: placebo	I1: sitagliptin 100mg o.d.+ pioglitazone 30 or 45mg/day   C1: placebo + pioglitazone 30 or 45mg/day	I1: sitagliptin 5mg b.i.d.   I2: sitagliptin 12,5mg b.i.d.   I3: sitagliptin 25mg b.i.d.   I4: sitagliptin 50mg b.i.d.   C1: placebo   C2: glipizide uptitr 5‐20 mg	I1: sitagliptin 100mg o.d. + metformin >= 1500mg/day   C1: rosiglitazone 8mg o.d. + metformin >= 1500mg/day   C2: placebo + metformin >=1500 mg/day
[n] (I1/ I2 / C1 / total)	I1: 205   I2: 206   C1: 110   Total: 521	I1: 175   C1: 178   Total: 353	I1: 125   I2: 123   I3: 123   I4: 124   C1: 125   C2: 123   Total: 743	I1: 94   C1: 87   C2: 92   Total: 273
Sex [n,%]	I1: female 95 (46.3); male 110 (53.7)   I2: female 102 (49.5); male 104 (50.5)   C1: female 41 (37.3); male 69 (62.7)	I1: female 82 (46.9); male 95 (53.1)   C1: female 75 (42.1); male 103 (57.9)	I1: female 63 (50.4); male 62 (49.6)   I2: female 64 (52.0); male 59 (48.0)   I3: female 52 (42.3); male 71 (57.7)   I4: female 59 (47.6); male 65 (52.4)   C1: female 47 (37.6); male 78 (62.4)   C2: female 53 (43.1); male 70 (56.9)	I1: female 42 (45); male 52 (55)   C1: female 32 (37); male 55 (63)   C2: female 38 (41); male 54 (59)
Age [years] mean (SD)	I1: 54.5 (10.0)   I2: 55.4 (9.2)   C1: 55.5 (10.1)	I1: 55.6 (10.4)   C1: 56.9 (11.1)	I1: 55.1 (9.5)   I2: 56.2 (9.0)   I3: 55.6 (9.0)   I4: 55.1 (9.8)   C1: 55.3 (9.7)   C2: 54.7 (10.7)	I1: 55.2 (9.8)   C1: 54.8 (10.5)   C2: 55.3 (9.3)
Ethnic groups [%]	I1: white 69.3; black 7.8; hispanic 18.0; asian 3.9; other 1.0   I2: white 70.9; black 5.3; hispanic 18.9; asian 3.4; other 1.5   C1: white 61.8; black 10.9; hispanic 20.0; asian 4.5; other 2.7	I1: white 72.6; hispanic 12.0; black 6.3; asian 5.7; other 3.4   C1: white 72.5; hispanic 12.4; black 6.7; asian 2.8; other 5.6	I1: asian 5.6; black 6.4; multi‐racial 6.4; white 68.8; other 12.8   I2: asian 4.9; black 4.9; multi‐racial 5.7; white 63.4; other 21.1   I3: asian 4.9; black 8.9; multi‐racial 6.5; white 61; other 18.7   I4: asian 2.4; black 4.8; multi‐racial 7.3; white 69.4; other 16.1   C1: asian 2.4; black 8.0; multi‐racial 7.2; white 66.4; other 16.0   C2:	I1: caucasian 61; asian 38; others 1   C1: caucasian 59; asian 38; others 3   C2: caucasian 61; asian 39; others 0
Duration of disease [years] mean (SD)	I1: 4.5 (4.3)   I2: 4.5 (3.9)   C1: 4.7 (5.0)	I1: 6.1 (5.4)   C1: 6.1 (5.7)	I1: 4.3 (4.1)   I2: 4.9 (5.0)   I3: 5.0 (5.2)   I4: 4.2 (4.0)   C1: 4.8 (4.7)   C2: 4.7 (4.2)	I1: 4.9 (3.5)   C1: 4.6 (4.0)   C2: 5.4 (3.7)
Body mass index [kg/m2] mean (SD)	I1: 31.8 (5.3)   I2: 32.0 (5.3)   C1: 32.5 (5.2)	I1: 32.0 (5.2)   C1: 31.0 (5.0	I1: 30.8 (5.1)   I2: 30.5 (5.0)   I3: 31.4 (6.9)   I4: 30.4 (4.9)   C1: 31.6 (5.8)   C2: 30.6 (5.3)	I1: 30.3 (4.7)   C1: 30.4 (5.5)   C2: 30.0 (4.5)
Pharmaco‐naive patients [n,%]	I1: 87 (42.4)   I2: 86 (41.7)   C1: 40 (36.4)	I1: 14 (8.0)   C1: 20 (11.3) *	I1: ?   I2: ?   I3: ?   I4: ?   C1: ?   C2: ?	I1: 0   C1: 0   C2: 0
HbA1c [%] mean (SD)	I1: 8.0 (0.8)   I2: 8.1 (0.9)   C1: 8.0 (0.9)	I1: 8.1 (0.8) §   C1: 8.0 (0.8)	I1: 7.9 (1.0)   I2: 7.9 (0.9)   I3: 7.9 (0.9)   I4: 7.8 (1.0)   C1: 7.9 (1.0)   C2: 7.9 (1.0)	I1: 7.8 (1.0)   C1: 7.7 (0.8)   C2: 7.7 (0.9)
Notes		* n = 177   § n = 174
Symbols & abbreviations:   Y = yes; N = no; ? = unclear   I = intervention; C = control

Characteristic	Ahren 2004	Ahren 2005	Bolli 2008	Bosi 2007	Dejager 2007	Fonseca 2007	Garber 2007
Intervention 1 (I1) / intervention 2 (I2) / control 1 (C1)	I1: vildagliptin 50mg q.d.+ metformin 1500‐ 3000mg/day   C1: placebo + metformin 1500‐ 3000mg/day	I1: vildagliptin 50mg o.d.+ metformin 1500‐ 3000mg/day   C1: placebo + metformin 1500‐ 3000mg/day	I1: vildagliptin 100 mg/day + metformin >= 1500 mg/day   C1: pioglitazone 30 mg/day + metformin >= 1500 mg/day	I1: vildagliptin 50mg o.d.+ metformin >= 1500mg/day   I2: vildagliptin 100mg o.d.+ metformin >= 1500mg/day   C1: placebo + metformin >=1500mg/day	I1: vildagliptin 50mg o.d.   I2: vildagliptin 50mg b.i.d..   I3: vildagliptin 100mg o.d.   C1: placebo	I1: vildagliptin 50mg b.i.d.+ insulin   C1: placebo + insulin	I1: vildagliptin 50mg o.d.+ pio 45mg o.d.   I2: vildagliptin100mg o.d.+ pio 45mg o.d.   C1: placebo + pio 45mg o.d.
[n] (I1/ I2 / C1 / total)	I1: 56   C1: 51   Total: 107 *	I1: 31   C1: 26   Total: 57 *	I1: 295   C1: 281   Total: 576	I1: 143   I2: 143   C1: 130   Total: 416 *	I1: 104   I2: 90   I3: 92   C1: 94   Total: 380 *	I1: 144   C1: 152   Total: 296	I1: 124   I2: 136   C1: 138   Total: 398 *
Sex [n,%]	I1: female 17 (30.4); male 39 (69.6)   C1: female 17 (33.3); male 34 (66.7) **	I1: female 9 (29); male 22 (71)   C1: female 6 (23); male 20 (77) **	I1: female 113 (38.3); male 182 (61.7)   C1: female 101 (35.9); male 180 (64.1)**	I1: female 61 (42.7); male 82 (57.3)   I2: female 55 (38.5); male 88 (61.5)   C1: female 61 (46.9); male 69 (53.1)	I1: female 61 (58.7); male 43 (41.3)   I2: female 48 (53.3); male 42 (46.7)   I3: female 43 (46.7); male 49 (53.3)   C1: female 49 (52.1); male 45 (47.9)	I1: female 75 (52.1); male 69 (47.9)   C1: female 69 (45.4); male 83 (54.6) **	I1: female 56 (45.2); male 68 (54.8)   I2: female 75 (55.1); male 61 (44.9)   C1: 68 (49.3); male 70 (50.7)
Age [years] mean (SD)	I1: 57.9 (10.0)   C1: 55.7 (11.0)	I1: 57.5 (9.2)   C1: 55.9 (10.0)	I1: 56.3 (9.3)   C1: 57.0 (9.7)	I1: 54.3 (9.7)   I2: 53.9 (9.5)   C1: 54.5 (10.3)	I1: 55.3 (11.4)   I2: 52.8 (9.6)   I3: 53.6 (10.8)   C1: 52.2 (11.2)	I1: 59.6 (10.3)   C1: 58.9 (10.8)	I1: 54.0 (8.2)   I2: 54.0 (9.2)   C1: 54.8 (10.6)
Ethnic groups [%]	I1: asian 1.8; caucasian 98.2 §   C1: ?	I1: N   C1: N	I1: caucasian 82.4; hispanic or latino 8.5; asian (non‐indian subcontinent)4.1; black 3.0; all others 2.0   C1: caucasian 81.9; hispanic or latino 10.3; asian (non‐indian subcontinent) 3.9; black 2.5; all others 1.4	I1: caucasian 74.1; hispanic or latino 16.8; black 6.3; all other 2.8   I2: caucasian 74.1; hispanic or latino 13.3; black 9.1; all other 3.5   C1: caucasian 73.1; hispanic or latino 18.5; black 6.9; all other 1.5	I1: caucasian 73.1; hispanic or latino 13.5; black 9.6; all other 3.8   I2: caucasian 73.3; hispanic or latino13.3; black 10.0; all other 33.4   I3: caucasian 76.1; hispanic or latino 15.2; black 4.3; all other 4.4   C1: cucasian 69.1; hispanic or latino 11.7; black 12.8; all other 6.4	I1: black 15.3; white 70.1; hispanic or latino 11.8; all others 2.8   C1: black 11.2; white 72.4; hispanic or latino 14.5; all others 2.0	I1: caucasian 83.9; hispanic or latino 9.7; black 4.8; all other 1.6   I2: caucasian 79.4; hispanic or latino 8.8; black 8.1; all other 3.7   C1: caucasian 78.3; hispanic or latino 7.2; black 9.4; all other 5.1
Duration of disease [years] mean (SD)	I1: 5.6 (4.2)   C1: 5.5 (3.7)	I1: 5.6 (4.2)   C1: 5.5 (3.7)	I1: 6.4 (4.9)   C1: 6.4 (5.2)	I1: 6.8 (5.5)   I2: 5.8 (4.7)   C1: 6.2 (5.3)	I1: 2.1 (3.6)   I2: 2.1 (3.3)   I3: 2.4 (4.2)   C1: 1.6 (2.5)	I1: 14.4 (8.6)   C1: 14.9 (8.4)	I1: 4.7 (4.3)   I2: 4.6 (4.8)   C1: 4.8 (4.6)
Body mass index [kg/m2] mean (SD)	I1: 29.4 (3.6)   C1: 30.2 (3.6)	I1: 29.3 (3.6)   C1: 29.8 (3.5)	I1: 32.2 (5.6)   C1: 32.1 (5.1)	I1: 32.1 (5.3)   I2: 32.9 (5.0)   C1: 33.2 (6.1)	I1: 32.9 (6.0)   I2: 33.3 (4.8)   I3: 32.4 (6.1)   C1: 32.6 (5.6)	I1: 33.3 (5.2)   C1: 32.9 (5.9)	I1: 32.6 (5.0)   I2: 32.2 (5.8)   C1: 32.3 (5.8)
Pharmaco‐naive patients [n,%]	I1: 0   C1: 0	I1: 0   C1: 0	I1: 0   C1: 0	I1: 0   I2: 0   C1: 0	I1: 100% §   I2: 100% §   I3: 100% §   C1: 100% §	I1: 0   C1: 0	I1: 0   I2: 0   C1: 0
HbA1c [%] mean (SD)	I1: 7.7 (0.6)   C1: 7.8 (0.7)	I1: 7.6 (0.6)   C1: 7.8 (0.7)	I1: 8.4 (1.0)   C1: 8.4 (0.9)	I1: 8.4 (0.9)   I2: 8.4 (1.0)   C1: 8.3 (0.9)	I1: 8.2 (0.8)   I2: 8.6 (0.8)   I3: 8.4 (0.8)   C1: 8.4 (0.8)	I1: 8.4 (1.0)   C1: 8.4 (1.1)	I1: 8.6 (1.0)   I2: 8.7 (1.2)   C1: 8.7 (1.2)
Notes	* Data ITT ( = randomized population) for the 12 ‐week core study   **sex: females calculated   § ethnic group calculated	* pat. completing 52 weeks with participation in all meal tests   ** sex calculated	** sex calculated for females	* primary ITT population for all baseline characteristics.   Randomised:   I1: 177   I2: 185   C1: 182   Total: 544	* primary ITT population for all baseline characteristics.   Randomised:   I1: 163   I2: 152   I3: 157   C1: 160   Total: 632   § Drug‐naive = no oral antidiabetic drug (OAD) for at least 12 weeks prior to screening and no OAD for > 3 consecutive months at any time in the past	** sex: females calculated	* primary ITT population for all baseline characteristics   Randomised:   I1: 147   I2: 158   C1: 158   Total: 463
Symbols & abbreviations:   Y = yes; N = no; ? = unclear   I = intervention; C = control

Characteristic	Characteristic	Pi‐Sunyer 2007	Pratley 2006	Ristic 2005	Rosenstock 2007a	Rosenstock 2007b	Scherbaum 2008	Schweizer 2007
Intervention 1 (I1) / intervention 2 (I2) / control 1 (C1)	I1: vildagliptin 50mg o.d.   I2: vildagliptin 50mg b.i.d..   I3: vildagliptin 100mg o.d.   C1: placebo	I1: vildagliptin 25mg b.i.d.   C1: placebo b.i.d.	I1: vildagliptin 25mg b.i.d.   I2: vildagliptin 25mg o.d.   I3: vildagliptin 50mg o.d.   I4: vildagliptin 100mg o.d.   C1: placebo	I1: vildagliptin 100mg o.d.   C1: rosiglitazone 8mg o.d.	I1: vildagliptin 100mg o.d.   I2: vilda 100mg o.d. + pio 30mg o.d.   I3: vilda 50 mg o.d.+ pio 15mg o.d.   C1: pioglitazone 30mg o.d.	I1: vildagliptin 100mg/ day   C1: metformin 2000 mg/day	I1: vildagliptin 50mg o.d.   C1: placebo	I1: vildagliptin 100 mg/day   C1: metformin 2000 mg/day
[n] (I1/ I2 / C1 / total)	I1: 88   I2: 83   I3: 91   C1: 92   Total: 354	I1: 70   C1: 28   Total: 98 *	I1: 51   I2: 54   I3: 53   I4: 63   C1: 58   Total: 279	I1: 459   C1: 238   Total: 697 *	I1: 154   I2: 148   I3: 144   C1: 161   Total: 607	I1: 526   C1: 254	I1:156   C1: 150   Total: 306	I1: 526   C1: 254   Total: 780
Sex [n,%]	I1: female 39 (44.3); male 49 (55.7)   I2: female 36 (43.4); male 47 (56.6)   I3: female 42 (46.2); male 49 (53.8)   C1: female 42 (45.7); male 50 (54.3)	I1: female 42 (60.0); male 28 (40.0)   C1: female 14 (50.0); male 14 (50.0)	I1: female 27 (52.9); male 24 (47.1)   I2: female 20 (37.0); male 34 (63.0)   I3: female 27 (50.9); male 26 (49.1)   I4: female 28 (44.4); male 35 (55.6)   C1: female 25 (43.1); male 33 (56.9)	I1: female 195 (42.5); male 264 (57.5)   C1: female 101(42.4); male 137 (57.6)	I1: female 56 (36.4); male 98 (63.6)   I2: female 62 (41.9); male 86 (58.1)   I3: female 60 (41.7); male 84 (58.3)   C1: female58 (36.0); male 103 (64.0)	I1: female 248 (47.1); male 278 (52.9)   C1: female 108 (42.5); male 146 (57.5)	I1: female 63 (40.4); male 93 (59.6)   C1: female 61 (40.6); male 89 (59.3)**	I1: female 248 (47.1); male 278 (52.9)   C1: female 108 (42.5); male 146 (57.5)
Age [years] mean (SD)	I1: 50.6 (10.4)   I2: 50.2 (12.7)   I3: 52.0 (11.7)   C1: 52.0 (12.0)	I1:56.9 (9.4)   C1: 52.8 (10.0)	I1: 55.6 (10.9)   I2: 57.4 (10.2)   I3: 57.0 (10.2)   I4: 56.2 (10.1)   C1: 54.6 (10.6)	I1: 54.5 (11.7)   C1: 54.2 (11.6)	I1: 51.4 (10.8)   I2: 51.0 (11.3)   I3: 51.0 (11.0)   C1: 52.4 (10.3)	I1: 52.8 (11.7)   C1: 53.6 (10.2)	I1: 63.3 (10.2)   C1: 62.8 (11.0)	I1: 52.8 (11.7)   C1: 53.6 (10.2)
Ethnic groups [%]	I1: caucasian 54.5; hispanic or latino 18.2; asian 15.9 (indian subcontinent); asian 3.4 (non‐indian subcontinent); black 8.0   I2: caucasian 53.0; hispanic or latino 21.7; asian 18.1 (indian subcontinent); asian 1.2 (non‐indian   subcontinent); black 6.0   I3: caucasian 58.2; hispanic or latino 12.1; asian 16.5 (subcontinent); asian 1.1 (non‐indian subcontinent); black 12.1   C1: caucasian 51.1; hispanic or latino 18.5; asian 16.3 (indian subcontinent); asian 1.1 (non‐indian subcontinent); black 13.0	I1: black 2.9; caucasian 47.1; oriental 1.4; other 48.6   C1: black 0; caucasian 46.4; oriental 0; other 53.6	I1: caucasian 80.4   I2: caucasian 79.6   I3: caucasian 77.4   I4: caucasian 74.6   C1: caucasian 87.9	I1: caucasian 79.5; hispanic or latino 11.1; black 5.9; all other 3.5   C1: caucasian 79.8; hispanic or latino 12.2; black 4.6; all other 3.4	I1: asian 45.5; caucasian 39.0; hispanic or latino 11.0; all other 4.5   I2: asian 44.7; caucasian 37.8; hispanic or latino 15.5; all other 2.0   I3: asian 47.2; caucasian 36.1; hispanic or latino 10.4; all other 6.3   C1: asian 42.9; caucasian 44.1; hispanic or latino 8.7; all other 4.3	I1: caucasian 67.9; hispanic or latino 19.8; black 8.0; all other 4.3   C1: caucasian 69.7; hispanic or latino 21.7; black 5.1; all other 3.5	I1: caucasian 99.4; others 0.6   C1: caucasian 99.3; others 0.7	I1: caucasian 357 (67.9); hispanic or latino 104 (19.8); black41 (8.0); all other 23 (4.3)   C1: caucasian 177 (69.7); hispanic or latino 55 (21.7); black 13 (5.1); all other 9 (3.5)
Duration of disease [years] mean (SD)	I1: 1.8 (2.7)   I2: 2.4 (3.2)   I3: 2.1 (2.9)   C1: 2.5 (3.7)	I1: 4.6 (5.6)   C1: 3.5 (5.7)	I1: 3.28 (3.81)   I2: 3.10 (5.16)   I3. 2.71 (3.24)   I4: 3.03 (4.22)   C1: 2.28 (2.99)	I1: 2.3 (3.4)   C1: 2.7 (4.2)	I1: 1.9 (3.1)   I2: 2.0 (3.1)   I3: 2.0 (3.2)   C1: 2.2 (3.3)	I1: 1.05 (IQR 3.54)   C1: 1.03 (IQR 3.28)	I1: 2.5 (2.9)   C1: 2.7 (3.2)	I1: 1.05 (IQR 3.54)   C1: 1.03 (IQR 3.28)
Body mass index [kg/m2] mean (SD)	I1: 31.9 (5.4)   I2: 32.2 (6.0)   I3: 31.9 (5.0)   C1: 32.7 (6.4)	I1: 30.0 (4.5)   C1: 29.9 (4.1)	I1: 30.9 (5.23)   I2: 31.1 (3.89)   I3: 31.0 (3.90)   I4: 31.1 (4.01)   C1: 31.6 (4.41)	I1: 32.2 (5.7)   C1: 32.9 (6.0)	I1: 29.4 (5.8)   I2: 29.6 (5.8)   I3: 29.0 (5.4)   C1: 28.9 (5.5)	I1: 32.4 (5.7)   C1: 32.5 (5.7)	I1: 30.4 (4.9)   C1: 30.0 (4.9)	I1: 32.4 (5.7)   C1: 32.5 (5.7)
Pharmaco‐naive patients [n,%]	I1: 100% §   I2: 100% §   I3: 100% §   C1: 100% §	I1: 100%   C1: 100%	I1: ? §   I2: ? §   I3: ? §   I4: ? §   C1: ? §	I1: 100% §   C1: 100% §	I1: 100%   I2: 100%   I3: 100%   C1: 100% §	I1: 100% §   C1: 100% §	I1: 100% §   C1: 100% §	I1: 100% §   C1: 100% §
HbA1c [%] mean (SD)	I1: 8.4 (0.9)   I2: 8.4 (0.9)   I3: 8.3 (0.8)   C1: 8.5 (0.8)	I1: 8.0 (0.9)   C1: 8.1 (1.2)	I1: 7.64 (0.69)   I2: 7.73 (0.80)   I3: 7.70 (0.82)   I4: 7.64 (0.75)   C1: 7.76 (0.83)	I1: 8.7 (1.1)   C1: 8.7 (1.1)	I1: 8.6 (1.0)   I2: 8.8 (1.1)   I3: 8.8 (0.9)   C1: 8.7 (1.0)	I1: 8.7 (1.1)   C1: 8.7 (1.1)	I1: 6.7 (0.4)   C1: 6.8 (0.4)	I1: 8.7 (1.1)   C1: 8.7 (1.1)
Notes	§ drug‐naive = no oral antidiabetic drug (OAD) for at least 12 weeks prior to screening and no OAD for >3 consecutive months at any time in the past.	* ITT population for baseline characteristics.   Randomised:   I1: 72   C1: 28   Total: 100	§ "additional exclusion criteria were treatment with oral antidiabetic drugs or sodium channel blockers whitin the previous 12 weeks, combination oral antidiabetic therapy or insulin treatment within 6 months prior to study ".	* primary ITT population for baseline characteristics   Randomised:   I1: 519   C1: 267   Total: 786   § " these patients had received no pharmacologic treatment for at least 12 weeks before screening and antidiabetic agent for > 3 consecutive months at any time in the past anr were considered to be representative of a drug‐naive population.	§ patients enrolled in the study: while receiving no pharmacological treatment for at least 12 weeks prior to screening and no OAD for more than 3 consecutive months at any time in the past.	§ drug naive = patients who had taken no oral glucose‐lowering agents for at least 12 weeks prior to screening and no oral glucose ‐lowering agents for more than 3 consecutive months at any time in the past .	** sex calculated for females   § drug‐naive = patients who had taken no oral glucose‐lowering agents for at least 12 weeks prior to screening and no oral glucose‐lowering agents for more than three consecutive months at any time inthe past .	§ drug naive = patients who had taken no oral glucose‐lowering agents for at least 12 weeks prior to screening and no oral glucose‐lowering agents for more than three consecutive months at any time inthe past .
Symbols & abbreviations:   Y = yes; N = no; ? = unclear   I = intervention; C = control   IQR = interquartile range

Study	Dosage [mg]	ITT/HbA1c population	Mean change f. base.	SD
Ahren 2004	vilda 50 mg o.d. + metformin 1500‐3000 mg	56	‐0.4	1.5
Ahren 2004	placebo + metformin 1500‐3000 mg	51	‐0.5	1.4
Bolli 2008	vilda 50 mg b.d. + metformin >= 1500 mg/day	264	0.3	3.2
Bolli 2008	pio 30 mg o.d. + metformin >= 1500 mg/day	246	1.9	3.1
Bosi 2007	vilda 50 mg o.d. + metformin >= 1500 mg	143	‐0.4	3.6
Bosi 2007	vilda 100 mg o.d. + metformin >= 1500 mg	143	0.2	3.6
Bosi 2007	placebo + metformin >= 1500 mg	130	‐1	3.4
Dejager 2007	vilda 50 mg o.d.	104	‐1.8	4.1
Dejager 2007	vilda 50 mg b.i.d.	90	‐0.3	3.8
Dejager 2007	vilda 100 mg o.d.	92	‐0.8	3.8
Dejager 2007	placebo	94	‐1.4	3.9
Fonseca 2007	vilda 50 mg b.i.d. + insulin	140	1.3	3.5
Fonseca 2007	placebo + insulin	149	0.6	3.7
Garber 2007	vilda 50 mg o.d. + pioglitazone 45 mg	124
Garber 2007	vilda 100 mg o.d. + pioglitazone 45 mg	136
Garber 2007	placebo + pioglitazone 45 mg	138	1.4	3.5
Mimori 2006	vilda 10 mg b.i.d.
Mimori 2006	vilda 25 mg b.i.d.
Mimori 2006	vilda 50 mg b.i.d.
Mimori 2006	placebo
Pi‐Sunyer 2007	vilda 50 mg o.d.	84	‐0.4
Pi‐Sunyer 2007	vilda 50 mg b.i.d.	79	‐0.4
Pi‐Sunyer 2007	vilda 100 mg o.d.	89	‐0.4
Pi‐Sunyer 2007	placebo	88	‐1.4	3.8
Pratley 2006	vilda 25 mg b.i.d.	70
Pratley 2006	placebo b.i.d.	28
Ristic 2005	vilda 25 mg b.i.d.	51	0.06	2.4?
Ristic 2005	vilda 25 mg o.d.	54	‐0.55	2.4?
Ristic 2005	vilda 50 mg o.d.	53	0.04	2.4?
Ristic 2005	vilda 100 mg o.d.	63	‐0.07	2.5?
Ristic 2005	placebo	58	‐0.73	2.5?
Rosenstock 2007a	vilda 100 mg o.d.	459	‐0.3
Rosenstock 2007a	rosiglitazone 8 mg	238	1.6
Rosenstock 2007b	vilda 100 mg o.d.	150	0.2	3.7
Rosenstock 2007b	vilda 50 mg o.d. + pioglitazone 15 mg	139	1.4	3.5
Rosenstock 2007b	vilda 100 mg o.d. + pioglitazone 30 mg	146	2.1	3.6
Rosenstock 2007b	pioglitazone 30 mg	157	1.5	3.8
Scherbaum 2008	vilda 50 mg o.d.	153	‐0.5	3.7
Scherbaum 2008	placebo	149	‐0.2	3.7
Schweizer 2007	vilda 100 mg o.d.	511	0.3	4.5
Schweizer 2007	metformin 200 mg	249	‐1.9	4.7
Symbols & abbreviations:   ITT = intention‐to‐treat

study	HOMA beta: n	change from baseline	lower CI	upper CI	HOMA‐IR: n	change from baseline	lower CI	upper CI	comments
Aschner 2006:   100 mg o.d.	218	13.2	6.7	19.7
Aschner 2006:   200 mg o.d.	228	13.1	6.8	19.5
Aschner 2006:   placebo	235	0.3	‐6	6.5
Charbonnel 2006:   100 mg o.d. + metformin	418	19.5	12.9	26.2	418	0	‐0.6	0.6
Charbonnel 2006:   placebo + metformin	196	3.5	‐4.9	11.8	196	0	‐0.7	0.7
Goldstein 2007:   100 mg o.d.	147	10.8	4.8	16.9	147	‐0.2	‐0.8	0.4
Goldstein 2007:   50 mg + metformin 500 mg b.i.d.	166	31	25.3	36.7	166	‐0.8	‐1.4	‐0.2
Goldstein 2007:   50 mg + metformin 1000 mg b.i.d.	160	33	27.2	38.8	160	‐2.4	‐2.9	‐1.8
Goldstein 2007:   metformin 500 mg b.i.d.	159	11.1	5.3	16.9	159	‐0.7	‐1.3	‐0.1
Goldstein 2007:   metformin 1000 mg b.i.d.	154	14.3	8.4	20.3	154	‐1.3	‐1.9	‐0.7
Goldstein 2007:   placebo	139	3.7	‐2.5	9.9	139	0.3	‐0.3	1
Hanefeld 2007:   25 mg o.d.	104	10	2.4	17.6	104	‐0.1	‐1	0.8
Hanefeld 2007:   50 mg o.d.	100	10.6	2.8	18.3	100	‐0.3	‐1.2	0.7
Hanefeld 2007:   100 mg o.d.	97	11.1	3.2	18.9	97	‐0.5	‐1.5	0.4
Hanefeld 2007:   50 mg b.i.d.	101	13.8	6.1	21.6	101	‐0.2	‐1.2	0.7
Hanefeld 2007:   placebo	95	‐1.4	‐9.3	6.6	95	‐0.1	‐1.1	0.8
Hermansen 2007:   100 mg o.d.+ glimepiride +/ ‐ metformin	186	11.3	4.4	18.1
Hermansen 2007:   placebo + glimepiride +/ ‐ metformin	156	‐0.7	‐8.2	6.8
Nauck 2007:   100 mg o.d.+ metformin	368	3.6	‐4.1	11.3	368	‐0.1	‐0.5	0.4
Nauck 2007:   glipizide + metformin	387	14	6.5	21.5	388	0.2	‐0.3	0.6
Nonaka 2008:   100 mg o.d.	75	9.5	6.1	12.9	75	‐0.15	‐0.42	0.13
Nonaka 2008:   placebo	74	‐3.1	‐6.5	0.3	74	0.09	‐0.19	0.36
Raz 2007:   100 mg o.d.	168	12.1	6	18.3
Raz 2007:   200 mg o.d.	171	13	6.9	19.2
Raz 2007:   placebo	80	1	‐8	10
Rosenstock 2006:   100 mg o.d.+ pioglitazone	133	11.5	6	17	133	‐0.1	‐0.6	0.4
Rosenstock 2006:   placebo + pioglitazone	142	5.8	0.7	10.9	142	0.2	‐0.3	0.6
Scott 2007a:   5 mg b.i.d.	115	8.3	0.9	15.7	115	0.6	‐0.2	1.1
Scott 2007a:   12,5 mg b.i.d.	118	8.2	0.9	15.5	118	0	‐0.9	0.8
Scott 2007a:   25 mg b.i.d.	114	6.7	‐0.8	14.1	114	‐0.2	‐1.1	0.6
Scott 2007a:   50 mg b.i.d.	115	17.3	9.8	24.7	115	0.1	‐0.7	1
Scott 2007a:   glipizide	105	25.4	17.7	33.2	106	0.9	0	1.8
Scott 2007a:   placebo	112	‐0.6	‐8.1	6.9	113	0.3	‐0.6	1.1
Scott 2007b:   100 mg o.d + metformin	78	9.4	‐0.4	19.2	78	‐0.5	‐1.1	0.2
Scott 2007b:   rosiglitazone 8 mg o.d.+ metformin	71	8.4	‐1.9	18.7	71	‐2.1	‐2.8	‐1.4
Scott 2007b:   placebo	76	‐6.9	‐16.8	3	76	0.3	‐0.4	1

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Change in haemoglobin A1c from baseline to endpoint	11	6910	Mean Difference (IV, Fixed, 95% CI)	‐0.54 [‐0.58, ‐0.50]
1.1 Sitagliptin versus placebo	6	1714	Mean Difference (IV, Fixed, 95% CI)	‐0.77 [‐0.85, ‐0.68]
1.2 Sitagliptin versus another single hypoglycaemic agent	2	592	Mean Difference (IV, Fixed, 95% CI)	0.33 [0.18, 0.48]
1.3 Sitagliptin combination therapy versus another hypoglycaemic combination	6	2890	Mean Difference (IV, Fixed, 95% CI)	‐0.40 [‐0.47, ‐0.33]
1.4 Sitagliptin versus placebo (12 weeks)	3	605	Mean Difference (IV, Fixed, 95% CI)	‐0.79 [‐0.90, ‐0.67]
1.5 Sitagliptin versus placebo (18 to 52 weeks)	3	1109	Mean Difference (IV, Fixed, 95% CI)	‐0.75 [‐0.86, ‐0.63]
2 Adverse events [n]	11	12416	Risk Ratio (M‐H, Fixed, 95% CI)	1.15 [1.02, 1.31]
2.1 Discontinuation due to adverse events	11	4414	Risk Ratio (M‐H, Fixed, 95% CI)	1.05 [0.77, 1.43]
2.2 Serious adverse events	11	4413	Risk Ratio (M‐H, Fixed, 95% CI)	0.97 [0.75, 1.27]
2.3 All‐cause infections	8	3589	Risk Ratio (M‐H, Fixed, 95% CI)	1.29 [1.09, 1.52]
3 Change in body weight from baseline to endpoint	4	1259	Mean Difference (IV, Fixed, 95% CI)	0.66 [0.37, 0.94]
3.1 Sitagliptin versus placebo	3	1109	Mean Difference (IV, Fixed, 95% CI)	0.69 [0.32, 1.06]
3.2 Sitagliptin versus another single hypoglycaemic agent	1	150	Mean Difference (IV, Fixed, 95% CI)	0.6 [0.13, 1.07]

Methods	DURATION OF INTERVENTION:   12 weeks   DURATION OF FOLLOW‐UP:   12 weeks   RUN‐IN PERIOD:   4‐week run‐in period in which patients received placebo while maintaining their previous metformin regimen   LANGUAGE OF PUBLICATION:   English
Participants	WHO PARTCIPATED:   patients with type 2 diabetes continuing a stable dosage of metformin (1500–3000 mg/day)   INCLUSION CRITERIA:   male or infertile female patients aged >=30 years diagnosed with type 2 diabetes at least 6 months before enrolment and treated with a stable dosage of metformin for >=3 months were included; prerandomization HbA1c while on metformin monotherapy was required to be between 7.0 and 9.5% (inclusive), and baseline BMI was required to be between 20 and 35 kg/m2 (inclusive)   EXCLUSION CRITERIA:   history of type 1 or secondary forms of diabetes, significant diabetes complications, clinically significant cardiovascular abnormalities, liver disease, acromegaly,   asthma, major skin allergies, or major gastrointestinal surgery; patients with fasting triglyceride levels >5.1 mmol/L or fasting plasma glucose (FPG) <6.1 or >=13.3   mmol/L were excluded, as were those treated with any drugs considered possibly able to affect results or their interpretation   DIAGNOSTIC CRITERIA:   nr   CO‐MORBIDITIES:   nr   CO‐MEDICATIONS:   nr
Interventions	NUMBER OF STUDY CENTRES:   nr   COUNTRY/ LOCATION:   multinational (2 countries)   SETTING:   nr   INTERVENTION (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   vildagliptin 50 mg o.d. (add‐on to metformin therapy)   CONTROL (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   placebo (add‐on to metformin therapy)   TREATMENT BEFORE STUDY:   patients with type 2 diabetes continuing a stable dosage of metformin (1500–3000 mg/day)   TITRATION PERIOD:   nr
Outcomes	see table "outcome data" under "Additional tables"
Notes	STATED AIM OF STUDY:   "to assess the 12‐ and 52‐week efficacy of the dipeptidyl peptidase IV inhibitor LAF237 (vildagliptin) versus placebo in patients with type 2 diabetes continuing metformin treatment the 12‐week core study was followed by a 40‐week extension in those patients completing the core study and agreeing, together with the investigator,   to participate"
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Unclear risk	Quote: "randomized ... trial"
Allocation concealment?	Unclear risk	not stated
Blinding?   HbA1c	Low risk	Quote: "double‐blind ... trial"
Incomplete outcome data addressed?   HbA1c	Low risk	Quote: "the primary efficacy variable ... in the intent‐to‐treat (ITT) population, with the last observation carried forward "
Free of selective reporting?	Low risk	none
Free of other bias?	Low risk	none

Methods	DURATION OF INTERVENTION:   24 weeks   DURATION OF FOLLOW‐UP:   24 weeks plus long‐term treatment period   RUN‐IN PERIOD:   patients with an A1C   of 7–10% and not on an OHA for >=8 weeks were eligible to directly enter a 2‐week single‐blind placebo run‐in period; patients with A1C >10% and not on an OHA entered a run‐in period of up to 6 weeks; patients with an A1C of 6–10% and on an OHA discontinued the agent and entered a wash‐out period of 6–10 weeks (8–12 weeks for those on thiazolidinediones); if A1C was 7–10% after the wash‐out period, patients were eligible to enter the placebo run‐in period   LANGUAGE OF PUBLICATION:   English
Participants	WHO PARTCIPATED:   patients with inadequately controlled type 2 diabetes   INCLUSION CRITERIA:   patients, 18–75 years of age, on and not on an OHA were eligible.   EXCLUSION CRITERIA:   patients with type 1 diabetes, unstable cardiac disease, significant renal impairment (creatinine clearance <50 ml/min), or elevated (more than twofold the upper limit of normal) alanine aminotransferase, aspartate aminotransferase, or creatine phosphokinase   DIAGNOSTIC CRITERIA:   nr   CO‐MORBIDITIES:   nr   CO‐MEDICATIONS:   nr
Interventions	NUMBER OF STUDY CENTRES:   nr   COUNTRY/ LOCATION:   multinational (16 countries)   SETTING:   nr   INTERVENTION (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   sitagliptin 100 or 200 mg o.d.   CONTROL (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   placebo   TREATMENT BEFORE STUDY:   on and not on an OHA   TITRATION PERIOD:   nr
Outcomes	see table "outcome data" under "Additional tables"
Notes	"STATED AIM OF STUDY:   to explore tolerability and potential dose‐dependent efficacy in patients with inadequately controlled type 2 diabetes"
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Unclear risk	Quote: "randomized ... study"
Allocation concealment?	Unclear risk	not stated
Blinding?   HbA1c	Low risk	Quote: "double‐blind ... study"; "laboratory measurements and ECGs were performed by technicians blinded to treatment groups"
Incomplete outcome data addressed?   HbA1c	Low risk	Quote: "efficacy analyses were based on the all‐patients treated population ... missing data were handled using the last observation carried forward method"
Free of selective reporting?	Low risk	none
Free of other bias?	Low risk	none

Methods	DURATION OF INTERVENTION:   24 weeks   DURATION OF FOLLOW‐UP:   24 weeks   RUN‐IN PERIOD:   none   LANGUAGE OF PUBLICATION:   English
Participants	WHO PARTCIPATED:   patients with type 2 diabetes inadequately controlled with metformin monotherapy   INCLUSION CRITERIA:   patients who were diagnosed with T2DM and had A1C of 7.5–11.0% at the screening visit while receiving a stable dose of metformin >=1500 mg/day; male and female (non‐fertile or of childbearing potential using a medically approved birth control method) patients aged 18–77 years, inclusive, with a body mass index (BMI) of 22–45 kg/m2, inclusive, and with FPG of <15 mmol/l   EXCLUSION CRITERIA:   history of type 1 or secondary formsof diabetes, acute metabolic diabetic complications, myocardial infarction, unstable angina or coronary artery bypass surgery within the previous 6 months; congestive heart failure (New York Heart Association [NYHA] classes I–IV) and liver disease such as cirrhosis or chronic active hepatitis also precluded participation; patients with any of the following laboratory abnormalities were also excluded: ALT or AST greater than 2.5 times the upper limit of normal (ULN), direct bilirubin >1.3 times the ULN, serum creatinine levels >132 mmol/L (males) or >125 mmol/L (females), clinically significant abnormal thyroid‐stimulating hormone or fasting triglycerides (TG) >7.9 mmol/L   DIAGNOSTIC CRITERIA:   nr   CO‐MORBIDITIES:   nr   CO‐MEDICATIONS:   nr
Interventions	NUMBER OF STUDY CENTRES:   118   COUNTRY/ LOCATION:   multinational (9 countries)   SETTING:   nr   INTERVENTION (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   vildagliptin 100 mg daily, given as two equally divided doses (add‐on to metformin therapy)   CONTROL (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   pioglitazone 30 mg o.d. (add‐on to metformin therapy)   TREATMENT BEFORE STUDY:   stable dose of metformin >=1500 mg/day   TITRATION PERIOD:   nr
Outcomes	see table "outcome data" under "Additional tables"
Notes	"STATED AIM OF STUDY:   to compare the efficacy and tolerability of vildagliptin and pioglitazone in patients with type 2 diabetes mellitus inadequately controlled with prior metformin monotherapy"
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Unclear risk	Quote: "randomized ... study"
Allocation concealment?	Unclear risk	not stated
Blinding?   HbA1c	Low risk	Quote: "double‐blind ... study"
Incomplete outcome data addressed?   HbA1c	Unclear risk	Quote: "the primary efficacy variable ... in the per protocol (PP) population using last observation carried forward for patients who discontinued early"; no intention‐to‐treat analysis
Free of selective reporting?	Low risk	none
Free of other bias?	Low risk	none

Methods	DURATION OF INTERVENTION:   24 weeks   DURATION OF FOLLOW‐UP:   24 weeks   RUN‐IN PERIOD:   none   LANGUAGE OF PUBLICATION:   English
Participants	WHO PARTCIPATED:   patients with type 2 diabetes inadequately controlled with metformin monotherapy   INCLUSION CRITERIA:   patients with type 2 diabetes who had been treated with metformin monotherapy for at least three months and who had been on a stable dose of >=1500 mg daily for a minimum of 4 weeks before visit 1; A1c in the range of 7.5‐11.0% at screening; male and female patients (nonfertile or of childbearing potential using a medically approved birth control method) aged 18‐78 years, inclusive, with a BMI in the range of 22‐45 kg/m2, inclusive, and with FPG <15 mmol/l   EXCLUSION CRITERIA:   history of type 1 or secondary forms of diabetes, acute metabolic diabetes complications within the last past 6 months, congestive heart failure requiring pharmacologic treatment, myocardial infarction, unstable angina, or coronary bypass surgery within the previous 6 months; liver disease such as cirrhosis or chronic active hepatitis; renal disease or renal dysfunction as suggested by elevated serum creatinine levels >=132 mmol/L for male and >=123 mmol/L for female participants   DIAGNOSTIC CRITERIA:   nr   CO‐MORBIDITIES:   nr   CO‐MEDICATIONS:   nr
Interventions	NUMBER OF STUDY CENTRES:   109   COUNTRY/ LOCATION:   multinational (4 countries)   SETTING:   nr   INTERVENTION (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   vildagliptin 50 mg o.d.   vildagliptin 50 mg b.i.d.   (add‐on to metformin therapy)   CONTROL (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   placebo   (add‐on to metformin therapy)   TREATMENT BEFORE STUDY:   metformin monotherapy, stable dose of >=1500 mg/day   TITRATION PERIOD:   nr
Outcomes	see table "outcome data" under "Additional tables"
Notes	STATED AIM OF STUDY:   "to examine the effects of vildagliptin in patients with type 2 diabetes inadequately controlled with metformin monotherapy"
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Unclear risk	Quote: "randomized ... study"
Allocation concealment?	Unclear risk	not stated
Blinding?   HbA1c	Low risk	Quote: "double‐blind ... study"
Incomplete outcome data addressed?   HbA1c	Low risk	Quote: "the primary efficacy variable ... using last observation carried forward for patients who discontinued early"; "the primary intent‐to‐treat (ITT) population ... was prespecified as the main efficacy population"
Free of selective reporting?	Low risk	none
Free of other bias?	Low risk	none

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Change in haemoglobin A1c from baseline to endpoint	14	6308	Mean Difference (IV, Fixed, 95% CI)	‐0.27 [‐0.29, ‐0.26]
1.1 Vildagliptin versus placebo	6	1139	Mean Difference (IV, Fixed, 95% CI)	‐0.32 [‐0.34, ‐0.30]
1.2 Vildagliptin versus another single hypoglycaemic agent	3	1764	Mean Difference (IV, Fixed, 95% CI)	0.30 [0.14, 0.46]
1.3 Vildagliptin combination therapy versus another hypoglycaemic combination	6	1756	Mean Difference (IV, Fixed, 95% CI)	‐0.05 [‐0.10, 0.01]
1.4 Vildagliptin versus placebo (12 weeks)	3	361	Mean Difference (IV, Fixed, 95% CI)	‐0.90 [‐1.05, ‐0.75]
1.5 Vildagliptin versus placebo (18 to 52 weeks)	4	1288	Mean Difference (IV, Fixed, 95% CI)	‐0.26 [‐0.28, ‐0.24]
2 Adverse events [n]	13	11562	Risk Ratio (M‐H, Fixed, 95% CI)	0.99 [0.87, 1.14]
2.1 Discontinuation due to adverse events	13	4543	Risk Ratio (M‐H, Fixed, 95% CI)	1.02 [0.75, 1.38]
2.2 Serious adverse events	11	3446	Risk Ratio (M‐H, Fixed, 95% CI)	0.87 [0.64, 1.17]
2.3 All‐cause infections	10	3573	Risk Ratio (M‐H, Fixed, 95% CI)	1.04 [0.87, 1.24]
3 Change in body weight from baseline to endpoint	5	1349	Mean Difference (IV, Fixed, 95% CI)	1.32 [1.02, 1.63]
3.1 Vildagliptin versus placebo	3	484	Mean Difference (IV, Fixed, 95% CI)	0.76 [0.19, 1.32]
3.2 Vildagliptin versus another single hypoglycaemic agent	2	865	Mean Difference (IV, Fixed, 95% CI)	1.55 [1.19, 1.91]

Methods	DURATION OF INTERVENTION:   52 weeks   DURATION OF FOLLOW‐UP:   52 weeks   RUN‐IN PERIOD:   patients who were already on metformin >=1500 mg/day and had an HbA1c >=6.5 and <=10% directly entered a 2‐week placebo run‐in period and were eligible to be randomized; patients not currenty on an OHA, patients on an OHA other than metformin monotherapy at a dose >=1500 mg/day or patients on metformin in combination with another OHA entered a metformin monotherapy treatment titration and dose‐stable period of at least 8 weeks; patients with an HbA1 >=6.5 and <=10% after the metformin dose‐stable period entered a 2‐week single‐blind placebo run‐in period   LANGUAGE OF PUBLICATION:   English
Participants	WHO PARTCIPATED:   patients with type 2 diabetes and inadequate glycaemic control on metformin monotherapy   INCLUSION CRITERIA:   men and women (age 18‐78 years) with type 2 diabetes who were not currently on an OHA, were taking any OHA in monotherapy or were taking metformin in combination with another OHA were potentially eligible to participate in the study if they all met screening criteria   EXCLUSION CRITERIA:   history of type 1 diabetes, insulin use within 8 weeks of screening, renal function impairment inconsistent with the use of metformin or a FPG (or a fasting fingerstick   glucose) at or just prior to randomization >15.0 mmol/L (270 mg/dl); other treatments for hyperglycaemia were prohibited during the study; concurrent lipid lowering and antihypertensive medications, thyroid medications, hormone replacement therapy and birth control medications were allowed but were expected to remain at stable doses   DIAGNOSTIC CRITERIA:   nr   CO‐MORBIDITIES:   nr   CO‐MEDICATIONS:   nr
Interventions	NUMBER OF STUDY CENTRES:   nr   COUNTRY/ LOCATION:   multinational (34 countries)   SETTING:   nr   INTERVENTION (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   addition of sitagliptin 100 mg o.d.   CONTROL (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   addition of glipizide o.d.   (at an initial dose of 5 mg/day)   TREATMENT BEFORE STUDY:   not currently on an OHA; taking any OHA in monotherapy or taking metformin in combination with another OHA   TITRATION PERIOD:   after the starting dose of 5 mg/day, glipizide was uptitrated according to protocol‐specified criteria to a potential maximum dose of 20 mg/day
Outcomes	see table "outcome data" under "Additional tables"
Notes	STATED AIM OF STUDY:   "to compare the glycaemic efficacy and safety of the addition of sitagliptin with that of a standard sulfonylurea agent, glipizide"
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Unclear risk	Quote: "randomized ... study; "randomized in a 1:1 ratio"
Allocation concealment?	Unclear risk	not stated
Blinding?   HbA1c	Low risk	Quote: "double‐blind ... study"
Incomplete outcome data addressed?   HbA1c	Low risk	Quote: "the primary efficacy analysis assessed ... using a per‐protocol (PP) approach; "additional efficacy analyses were   based on the all patients‐treated (APT) population ... missing values in the APT analysis were handled by the last observation   carried forward approach"
Free of selective reporting?	Low risk	none
Free of other bias?	Unclear risk	disparate and high attrition rates

Methods	DURATION OF INTERVENTION:   18 weeks   DURATION OF FOLLOW‐UP:   18 weeks + ongoing study   RUN‐IN PERIOD:   patients who entered the study on OHA therapy had the agent(s) discontinued and underwent a wash‐off and diet and exercise run‐in period of up to 12 weeks, based upon their prior therapy and HbA1c at study entry; patients not on an OHA (for >=8 weeks prior to screening visit) at study entry who met randomisation HbA1c criteria directly entered the 2‐week, single‐blind placebo run‐in period; patients whose HbA1c was >=7% and <=10% and who had adequate compliance (>=75%) during the single‐blind run‐in period were eligible to be randomised   LANGUAGE OF PUBLICATION:   English
Participants	WHO PARTCIPATED:   patients with type 2 diabetes mellitus and inadequate glycaemic control on exercise and diet   INCLUSION CRITERIA:   men and women with type 2 diabetes mellitus, 18–75 years of age, were recruited; patients not currently on oral antihyperglycaemic agent (OHA) therapy and patients on OHA monotherapy (or dual oral combination therapy in low doses) who could be taken off their OHA(s) during the run‐in period   EXCLUSION CRITERIA:   type 1 diabetes, insulin therapy, significant hepatic or renal disease, hepatic transaminase or creatine phosphokinase (CK) levels >=2 times the upper limit of normal, FPG >15 mmol/L (270 mg/dl) and BMI <20 kg/m2 or >43 kg/m2   DIAGNOSTIC CRITERIA:   nr   CO‐MORBIDITIES:   nr   CO‐MEDICATIONS:   nr
Interventions	NUMBER OF STUDY CENTRES:   nr   COUNTRY/ LOCATION:   multinational   SETTING:   nr   INTERVENTION (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   sitagliptin 100 mg and 200 mg o.d.   CONTROL (ROUTE, TOTAL DOSE/DAY, FREQUENCY):   placebo   [placebo, sitagliptin 100 mg and 200 mg in a 1:2:2 ratio]   TREATMENT BEFORE STUDY:   on or not on an OHA   TITRATION PERIOD:   nr
Outcomes	see table "outcome data" under "Additional tables"
Notes	STATED AIM OF STUDY:   "to assess the safety and efficacy of once‐daily sitagliptin 100 mg and 200 mg in patients with type 2 diabetes mellitus with inadequate glycaemic control on diet and exercise"; "the current report presents the initial 18‐week, placebo‐controlled study period; patients completing this period were eligible to enter an active‐controlled, double‐blind period, which was ongoing at the time of this report and which will be the subject of a later publication"
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Unclear risk	Quote: "randomized ... study"
Allocation concealment?	Unclear risk	not stated
Blinding?   HbA1c	Low risk	Quote: "double‐blind ... study"
Incomplete outcome data addressed?   HbA1c	Low risk	Quote: "all efficacy analyses were based on the all‐patients‐treated (APT) cohort ... missing data were handled by using the last observation carried forward method"
Free of selective reporting?	Low risk	none
Free of other bias?	Unclear risk	disparate attrition rates

PERMALINK

Dipeptidyl peptidase‐4 (DPP‐4) inhibitors for type 2 diabetes mellitus

Bernd Richter

Elizabeth Bandeira‐Echtler

Karla Bergerhoff

Christian Lerch

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Type 1 diabetes mellitus

Type 2 diabetes mellitus

Description of the intervention

Adverse effects of the intervention

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Covariates, effect modifiers and confounders

Timing of outcome measurement

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Dealing with duplicate publications

Assessment of risk of bias in included studies

Measures of treatment effect

Dichotomous data

Continuous data

Time‐to‐event data

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

Results

Description of studies

Results of the search

1.

Interrater agreement

Included studies

Interventions and comparisons

Sitagliptin

Comparisons with placebo

Comparisons with single hypoglycaemic agents

Comparisons of combination therapies

Vildagliptin

Comparisons with placebo

Comparisons with single hypoglycaemic agents

Comparisons of combination therapies

Number of study centres, countries, location and setting

Treatment before study

Duration of the intervention, run‐in period

Participants, inclusion and exclusion criteria, diagnostic criteria, co‐morbidities and medications

Primary outcomes

Secondary outcomes, additional/other outcomes

Excluded studies

Risk of bias in included studies

1. Study populations.

2.

3.

Allocation