SEARCH

SEARCH BY CITATION

Keywords:

  • cardiovascular;
  • GLP-1 receptor agonists;
  • type 2 diabetes

Summary

  1. Top of page
  2. Summary
  3. What is known and objective
  4. Methods
  5. Results
  6. Discussion
  7. What is new and conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

What is known and objective

Glucagon-like peptide-1 (GLP-1) receptor agonists are a new class of antidiabetic drugs. Their wider use for the treatment of patients with type 2 diabetes mellitus has led to concerns about its cardiovascular effects. However, the robustness of data leading to those concerns is unclear. The purpose of this study is to systematically assess the robustness of the available evidence on the adverse cardiovascular effects of GLP-1 receptor agonists in patients with type 2 diabetes.

Methods

The Cochrane library, MEDLINE, EMBASE and www.clinicaltrials.gov were searched from inception through to 25 January 2013. Randomized controlled trials (RCTs) were selected if they compared GLP-1 receptor agonists with placebo or other drugs with a duration ≥12 weeks. Mantel–Haenszel odds ratio (MH-OR) of cardiovascular events with 95% confidence interval (CI) was estimated using a random effects model. Trial sequential analysis based on required information size with an assumption of plausible reductions in relative risk in the low-bias trials, 5% risk of a type I error and 20% risk of a type II error was used to explore the robustness of available evidence.

Results

Fifty-eight trials were included in the analysis (10 466 patients receiving GLP-1 receptor agonists and 7138 patients receiving comparators, respectively). Overall, the OR for cardiovascular events with GLP-1 receptor agonists was 0·52 (95% CI: 0·27–0·99) compared with placebo and 0·84 (95% CI: 0·52–1·36) with active controls. Trial sequential analyses showed that the actual accumulated sample size was only 11% (7445 of 65 212) and 13% (10 157 of 79 198) of the required information size for placebo-controlled trials and active-controlled trials, respectively. These results indicate that there is still insufficient evidence on cardiovascular events.

What is new and conclusion

GLP-1 receptor agonists do not seem to show any increased risk of cardiovascular events However, the available data from RCTs remain insufficient to confirm an absence of detrimental effect. More long-term trials and population-based studies are required to provide the necessary reassurance on the cardiovascular safety of GLP-1 receptor agonists.


What is known and objective

  1. Top of page
  2. Summary
  3. What is known and objective
  4. Methods
  5. Results
  6. Discussion
  7. What is new and conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

An increasing number of patients with type 2 diabetes mellitus are being treated with glucagon-like peptide-1 (GLP-1) receptor agonists, a new class of antidiabetic agents.[1, 2] GLP-1 receptor agonists are analogues of GLP-1, which enhance glucose-induced insulin secretion, inhibit glucagon secretion and slow down gastrointestinal motility.[3-5] Exenatide (Byetta; Eli Lilly & Co.) and liraglutide (Victoz; Novo Nordisk), the two earliest GLP-1 receptor agonists, were approved by the United States Food and Drug Administration (FDA) in 2005 and 2010, respectively.[6, 7] Several other GLP-1 receptor agonists including albiglutide, lixisenatide (AVE0010) and taspoglutide are currently in phase III clinical trials.

Following the recent debate on the cardiovascular safety of rosiglitazone,[8-11] more and more attention is being directed towards other oral antidiabetic drugs including GLP-1 receptor agonists. Recently, a meta-analysis of 36 RCTs showed no increased risk of cardiovascular events in patients treated with GLP-1 receptor agonists.[12] Meanwhile, a recent retrospective analysis of the LifeLink database indicated that exenatide twice-daily treatment was associated with a lower risk of cardiovascular events than with other glucose-lowering therapies in a real-world population.[13] However, the authors acknowledged the potential misclassification bias and the potential residual confounding due to the limitations of the database.[13] In addition, no further analyses were included taking into account of the duration of trials and diabetic conditions, as well as the treatment modalities. More importantly, the robustness of the findings has not been evaluated in any previous studies. Given the relatively small number of affected patients and events, the effect size observed in those analyses might be due to the play of chance rather than a true effect of treatments.[14, 15] Trial sequential analysis (TSA), an approach that combines conventional meta-analysis methodology with meta-analytic sample size considerations, has recently been used to adjust for random error as a result of repetitive testing as data accrue.[14-17] The addition of TSA can help in assessing the reliability of cumulative data and reduce the risk of false-positive results that is not addressed in conventional meta-analyses. Several meta-analyses applying TSA have recently been published, such as one focusing on antihypertensive drugs and risk of cancer and another focusing on intensive glycaemic control and cardiovascular mortality.[18, 19]

The aim of our study was to assess the cardiovascular effect of GLP-1 receptor agonists in patients with type 2 diabetes and to evaluate the strength of all available evidence from published or unpublished RCTs.

Methods

  1. Top of page
  2. Summary
  3. What is known and objective
  4. Methods
  5. Results
  6. Discussion
  7. What is new and conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

Data sources and searches

A comprehensive search of MEDLINE, EMBASE and The Cochrane library up to 25 January 2013 for any published RCTs on humans that are written in English was conducted, with the following keywords: exenatide, liraglutide, taspoglutide, albiglutide, lixisenatide, Victoza, Byetta, Bydureon, NN9924, GLP-1 and glucagon-like peptide 1. Completed but unpublished trials were identified from a search of www.clinicaltrials.gov website using the search criteria.

Study selection

Only trials in which GLP-1 receptor agonists were compared with placebo or other drugs in patients with type 2 diabetes and of duration of 12 weeks or more were included in the analysis. Any ongoing or completed studies but with no results on the website of clinicaltrial.gov were excluded from the analysis. The eligibility of the studies for inclusion criteria was assessed independently by two reviewers (SS. Wu and Y. Zhang).

Data extraction and quality assessment

Data were extracted using EpiData (version 3.02, reference) with respect to author, publication year, clinical trial registration number, sample size, trial duration, types of control, diabetes duration, patterns of treatment, age, level of HbA1c, reported cardiovascular events and information on methodology. The endpoint of interest was a composite of cardiovascular events, which consisted of major adverse cardiovascular events (MACEs)[20] defined by FDA plus heart failure. MACEs included cardiovascular death, myocardial infarction and stroke.[20] Any discrepancies with respect to data extraction were resolved by consensus between the two independent reviewers (SS. Wu and Y. Zhang) or by a senior investigator (F. Sun).

The quality of each selected study was assessed using the JADAD scale[21] on the basis of information regarding the method for randomization, allocation concealment, blinding procedures and reporting of withdrawals and dropouts. A trial was considered as low-bias trial if all of the four items were adequate. However, the JADAD score was not used as a selection criterion, but only for descriptive purposes.

Statistical analysis

Mantel–Haenszel odds ratio (MH-OR) with 95% confidence interval was calculated using a random effect model due to the impossibility of a reliable heterogeneity assessment of trials with small number of events. All reported P values were 2-sided. Statistical heterogeneity was assessed using the Q-statistics. Placebo-controlled trials and active comparator trials were analysed separately. Additional subgroup analyses for the two categories of trials were performed based upon trial duration (categorized as long-term duration: ≥52 weeks; medium-term duration: between 26 to 52 weeks; short-term duration: <26 weeks), diabetes duration (categorized as ≥10 years; 5–10 years; less than 5 years) and patterns of treatments (monotherapy vs. add-on therapy). Besides, a separate analysis for exenatide and liraglutide, respectively, was also performed. As studies may contain no cardiovascular event, the constant continuity correction method was used with addition of a correction factor of 0·5 to the number of events and non-events in both treatment groups.[22-24]

Additionally, TSA was used to evaluate the robustness of the available evidence. Trial sequential analysis is similar to interim analyses with monitoring boundaries used to determine whether a trial could be terminated early when a P value is sufficiently small to show the effect.[14-17] Trial sequential analysis depends on the quantification of the required information size. In this context, the smaller the required information size is, the more lenient the trial sequential monitoring boundaries are and hence the more lenient the criteria for statistical significance. Heterogeneity (I²)-adjusted required information size was calculated with a desire to maintain an overall 5% risk of a type I error, which is the standard in most meta-analyses and systematic reviews. And the required information size calculations were based on an assumption of plausible reductions in relative risk in the low-bias trials for placebo-controlled trials and active comparator trials, respectively, with a risk of a type II error of 20% (power of 80%).[15, 16] The trial sequential analysis adjusted 95% confidence intervals were calculated too.

Excel version 2010 was used to sort the data. All analyses were performed using TSA version 0.9 beta (www.ctu.dk/tsa), except the test for publication bias that was performed by stata version 10.0.

Results

  1. Top of page
  2. Summary
  3. What is known and objective
  4. Methods
  5. Results
  6. Discussion
  7. What is new and conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

Study characteristics

A total of 47 published studies met the selection criteria and 6 of which were further excluded from the analysis due to absence of data on cardiovascular events. Five studies that included two comparator arms were treated as two separate trials, and one study that includes three comparator arms was treated as three separate trials. Eight studies from the clinicaltrial.gov website were included. Two studies that included two comparator arms were treated as two separate trials each. Therefore, 58 trials were included in the final analysis (Fig. 1). The characteristics of the included trials are summarized in Table S1 (online only). Of the 58 trials, a total of 17 604 patients were included, where 10 466 patients received GLP-1 receptor agonists, whereas 7138 patients received other comparators. Exenatide and liraglutide were studied in 33 and 18 studies, respectively. Of the 58 trials, there were 29 placebo-controlled trials and 29 active comparator trials.

image

Figure 1. Flow chart of studies considered for inclusion, RCT = randomized controlled trial.

Download figure to PowerPoint

According to JADAD scale,[21] the methods used for randomization and allocation concealment were not clearly stated in some instances (74% and 52%, respectively). Conversely, the number of withdrawals and methods used for blinding were appropriately described in all cases (98% and 100%, respectively). There were 24 trials with low bias (Table S2; online only).

Risk of cardiovascular events

Table S2 lists the cardiovascular events reported in the 58 included trials, with 49 events in the GLP-1 receptor agonist group and 42 events in the control groups.

As shown in Table 1, the overall OR for cardiovascular events associated with GLP-1 receptor agonists was 0·52 (95% confidence interval [CI], 0·27–0·99, P = 0·047) compared with placebo (N = 29). After the exclusion of short-term trials, the OR changed to 0·55 (95% CI: 0·20–1·52, = 0·249) (N = 9). For the placebo-controlled trials (Table 1), the risk for cardiovascular events with GLP-1 receptor agonists was 0·51 (95% CI: 0·22–1·15, P = 0·105) in trials of short-term duration (N = 20), whereas the risk was 0·37 (95% CI:0·12–1·16,P = 0·088) and 2·62 (95% CI:0·24–29·13,P = 0·433) in trials with a medium-term duration (N = 7) and a long-term duration (N = 2), respectively. For trials with monotherapy (N = 5) and add-on therapy (N = 24), the ORs were 0·30 (95% CI:0·06–1·62,P = 0·163) and 0·57 (95% CI: 0·28–1·15, P = 0·116), respectively. The ORs, 95% CIs and P values of all subgroup analyses are listed in Table 1. Results showed no heterogeneity of effects across trials (P > 0·05).

Table 1. Risk of cardiovascular events in subgroup analyses of placebo-controlled trials
 No. of study(N)Q-statistics valueOdds ratio (95% CI)P value
Trial duration
<26 weeks201·680·51(0·22,1·15)0·105
26~52 weeks71·470·37(0·12,1·16)0·088
≥52 weeks21·072·62(0·24,29·13)0·433
Diabetes duration
Not reported31·090·30(0·05,1·93)0·206
<5 years50·090·42(0·07,2·45)0·336
5~10 years175·640·53(0·23,1·25)0·147
≥10 years41·880·84(0·17,3·99)0·822
Patterns of treatment
Monotherapy51·150·30(0·06,1·62)0·163
Add-on therapy247·820·57(0·28,1·15)0·116
Total299·440·52(0·27,0·99)0·047

Compared with placebo, TSA indicated that the heterogeneity-adjusted information size required to demonstrate or reject a 33% relative risk reduction by the low-bias trials was 65 212 patients. This is much more than the number of patients who were actually accrued (trial sequential analysis adjusted 95% CI: 0·04–7·21). None of the boundaries for benefit or harm was crossed, and futility boundaries could not be produced, indicating that there was too little evidence to conclude whether GLP-1 receptor agonists were beneficial, harmful or without any effect with respect to cardiovascular events when compared with placebo (Fig. 2).

image

Figure 2. Trial sequential analysis of cardiovascular events compared with placebo. Heterogeneity-adjusted required information size of 65, 212 participants calculated on the basis of proportion of cardiovascular events of 0·36% in control group, and relative risk reduction in 33%, α = 5%, β = 20%, and I2 = 0%. The actual accrued number of participants was 7445, 11% of the required information size. The full cumulative Z curve does not cross the trial sequential monitoring boundaries and futility boundaries for benefit or harm. Horizontal dotted lines illustrate the traditional level of statistical significance (P = 0·05).

Download figure to PowerPoint

When only active comparator trials were analysed (N = 29), the summary OR for cardiovascular events was 0·84 (95% CI: 0·52–1·36, P = 0·473) in the GLP-1 receptor agonists group. After the exclusion of short-term trials, the OR changed to 0·87 (95% CI: 0·50–1·49, P = 0·604) (N = 20). In subgroup analyses of the active comparator trials (Table 2), the risk associated with GLP-1 receptor agonists was 0·54 (95% CI: 0·22–1·32, P = 0·174) and 1·00 (95% CI: 0·57–1·77, P = 0·989) for trials with monotherapy (N = 10) and add-on therapy (N = 19), respectively. In trials of long duration (N = 6), the OR was 0·75 (95% CI: 0·37–1·53, P = 0·435). No heterogeneity effect was detected between these trials (P > 0·05).

Table 2. Risk of cardiovascular events in subgroup analyses of active comparator trials
 No. of study (N)Q-statistics valueOdds ratio (95% CI)P value
Trial duration
<26 weeks90·860·75(0·27,2·09)0·579
26~52 weeks146·591·05(0·45,2·44)0·906
≥52 weeks63·230·75 (0·37,1·53)0·435
Diabetes duration
Not reported20·232·16(0·30,15·45)0·442
<5 years81·010·50(0·14,1·79)0·285
5~10 years178·190·84(0·49,1·46)0·535
≥10 years20·031·29(0·10,16·25)0·844
Patterns of treatment
Monotherapy101·230·54(0·22,1·32)0·174
Add-on therapy198·541·00(0·57,1·77)0·989
Total2911·10·84(0·52,1·36)0·473

Trial sequential analysis showed a lack of sufficient evidence of benefit, harm or without any effect on cardiovascular events when compared with active controls (trial sequential analysis adjusted 95% CI: 0·12–5·95). Only 10, 157 (13%) of the required heterogeneity-adjusted 79, 198 patients required to detect a 22% relative risk reduction were actually accrued (Fig. 3).

image

Figure 3. Trial sequential analysis for cardiovascular events compared with active comparators. Heterogeneity-adjusted required information size of 79,198 participants calculated on the basis of proportion of cardiovascular events of 0·73% in control group, relative risk reduction of 22%, α = 5%, β = 20% and I2 = 0%. The actual accrued number of participants was 10 157, 13% of required information size. Full cumulative Z curve does not cross the trial sequential monitoring boundaries and futility boundaries for benefit or harm. Horizontal dotted lines illustrate the traditional level of statistical significance (P = 0·05).

Download figure to PowerPoint

Separate analyses for exenatide showed that the risk of cardiovascular events was 0·46 (95% CI: 0·19–1·10, P = 0·819) compared with placebo (N = 15) and 0·91 (95% CI: 0·43–1·93, P = 0·802) with active control (N = 18). For liraglutide, the OR was 0·58 (95% CI: 0·17–2·02, P = 0·390) and 0·79 (95% CI: 0·42–1·48, P = 0·464) for placebo-controlled (N = 8) and active comparator trials (N = 10), respectively.

Discussion

  1. Top of page
  2. Summary
  3. What is known and objective
  4. Methods
  5. Results
  6. Discussion
  7. What is new and conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

Increasing attention is being paid to cardiovascular safety of diabetes drugs since the publication of a meta-analysis by Nissen et al. in 2007.[8] As a new class of antidiabetic drugs, our conventional meta-analysis, with predefined subgroup analyses, suggests that GLP-1 receptor agonists do not increase the risk of cardiovascular events, in accord with two previously published meta-analyses[12, 25] and a retrospective cohort study.[13]

Several mechanisms may account for this absence of detrimental effect of GLP-1 receptor agonists on cardiovascular events. First, GLP-1 receptor agonists may lower the risk of cardiovascular events by reducing multiple cardiovascular risk factors including blood pressure, cholesterol level and body weight, besides improving glycaemic control. A recent meta-analysis of 17 RCTs showed that GLP-1 receptor agonists led to clinically significant weight loss relative to most active comparators.[26] Similar reduction in weight was also observed in a few other studies comparing GLP-1 receptor agonists with either placebo or insulin analogues.[27, 28] In addition to weight loss, recent RCTs suggested that GLP-1 receptor agonists have a positive effect on blood pressure (both systolic and diastolic) and lipid profile, particularly LDL-C levels[29-31] relative to other treatments. Secondly, the cardiovascular effect of GLP-1 receptor agonists may include a direct action on vascular endothelium and myocardium. This may lead to beneficial vasodilatation and improvement in cardiac function. In animal models, exenatide produced a significant reduction in carotid artery intima/media ratio (a surrogate marker of cardiovascular disease) and monocyte/macrophage accumulation in the arterial wall.[32, 33] Recent studies in subjects with impaired glucose tolerance or recent onset type 2 diabetes demonstrated that acute administration of exenatide following a high-fat meal improved endothelial function.[34] Reduction in myocardial infarction size and improvements in cardiac function have been seen in animal models,[35, 36] although there is limited evidence in patients with heart failure or myocardial infarction. Furthermore, a recent prespecified and independently adjudicated cardiovascular meta-analysis including 8 RCTs of linagliptin suggested a potential reduction in cardiovascular events when compared with pooled comparators.[37] Linagliptin may increase available GLP-1 level.

Although several meta-analyses have been conducted on the risk of cardiovascular events, none has evaluated the strength of the available evidence or taken account of the information size or the number of accrued events. The positive finding in comparison with placebo and the negative finding in comparison with active controls may both be due to lack of statistical power. However, absence of evidence is not equal to the evidence of absence.[38] In this context, the required information size calculated by TSA, given the significant and non-significant findings of a meta-analysis, can provide further information with respect to whether more studies are needed. The TSA included in our study suggests that sufficient evidence is not yet available for any definite conclusion. In particular, additional 60 and 70 thousand patients are required to draw more robust estimates with a risk of a type I error of 5% and a type II error of 20% in comparison with placebo and active comparators, respectively.

A major strength of our study is the inclusion of a substantially greater number of trials than earlier meta-analyses and updating of the results of some trials.[12, 25]

Some limitations are worthy of mention. First, only trials in English were included, which may lead to a potential publication bias. However, the Begg's tests did not detect any evidence of publication bias (Fig. 4, P = 0·420; Fig. 5, P = 0·837). Furthermore, most trials included in this review were not specifically designed to evaluate cardiovascular outcomes, with the risk of underdiagnosis. Many of the trials were small and short-term, resulting in few cardiovascular events. As MACE was a composite outcome, it might lead to double-counting, for example, when someone with MI goes on to develop heart failure. All of the above may reduce the statistical power and precision of the estimation. The EXenatide Study of Cardiovascular Event Lowering trial (EXSCEL, NCT01144338),[39] the liraglutide trial (LEADER, NCT01179048)[39] and the lixisenatide trial (ELIXA, NCT01147250),[39] which were recently registered in clinicaltrials.gov, are all prospective, long-term, adjudicated cardiovascular outcome studies. However, they are still in the patient recruitment phase. Finally, we did not have access to original data for any of these trials, making the statistically more powerful time-to-event analysis not possible.

image

Figure 4. Begg's funnel plot for the placebo-controlled trials with pseudo 95% confidence limits.

Download figure to PowerPoint

image

Figure 5. Begg's funnel plot for the active comparator trials with pseudo 95% confidence limits.

Download figure to PowerPoint

What is new and conclusion

  1. Top of page
  2. Summary
  3. What is known and objective
  4. Methods
  5. Results
  6. Discussion
  7. What is new and conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

GLP-1 receptor agonists do not seem to increase risk of cardiovascular events. However, data available from published RCTs remain insufficient to confirm this. Further specially designed long-term trials and population-based observational studies are required to answer this question.

Acknowledgements

  1. Top of page
  2. Summary
  3. What is known and objective
  4. Methods
  5. Results
  6. Discussion
  7. What is new and conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

S.Y. Zhan designed the study and revised the manuscript, T.P. Hong and Y. Chen edited the manuscript, S.S. Wu and Y. Zhang extracted the data, Z.R. Yang and F. Sun verified the data. S.S. Wu researched the data and wrote the manuscript. We wish to sincerely thank all of the study authors who promptly and graciously responded to our requests for information. No potential conflicts of interest relevant to this article were reported. No funding source was involved in the conception or development of the study.

References

  1. Top of page
  2. Summary
  3. What is known and objective
  4. Methods
  5. Results
  6. Discussion
  7. What is new and conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information
  • 1
    Hansen KB, Knop FK, Holst JJ et al. Treatment of type 2 diabetes with glucagon-like peptide-1 receptor agonists. Int J Clin Pract, 2009;63:11541160.
  • 2
    Goykhman S, Drincic A, Desmangles JC et al. Insulin Glargine: A review 8 years after its introduction. Expert Opin Pharmacother, 2009;10:705718.
  • 3
    Nauck MA, Heimesaat MM, Behle K et al. Effects of glucagon-like peptide 1 on counterregulatory hormone responses, cognitive functions, and insulin secretion during hyperinsulinemic, stepped hypoglycemic clamp experiments in healthy volunteers. J Clin Endocrinol Metab, 2002;87:12391246.
  • 4
    Vilsboll T, Krarup T, Madsbad S et al. Both GLP-l and GIP are insulinotropic at basal and postprandial glucose levels and contribute nearly equally to the incretin effect of a meal in healthy subjects. Regul Pept, 2003;114:115121.
  • 5
    Zander M, Madsbad S, Madsen JL et al. Effect of 6-week course of glucagon-like peptide 1 on glycaemic control, insulin sensitivity, and beta-cell function in type 2 diabetes: A parallel-group study. Lancet, 2002;359:824830.
  • 6
    Wajcberg E, Tavaria A. Exenaride: clinical aspects of the first incretin-mimetic for the treatment of type 2 diabetes mellitus. Expert Opin Pharmacother, 2009;10:135142.
  • 7
    Neumiller JJ, Campbell RK. Liraglutide: A once-daily incretin mimetic for the treatment of type 2 diabetes mellitus. Ann Pharmacother, 2009;43:14331444.
  • 8
    Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med, 2007;356:24572471.
  • 9
    Singh S, Loke YK, Furberg CD. Long-term risk of cardiovascular events with rosiglitazone: a meta-analysis. JAMA, 2007;298:11891195.
  • 10
    Lago RM, Singh PP, Nesto RW. Congestive heart failure and cardiovascular death in patients with prediabetes and type 2 diabetes given thiazolidinediones: a meta-analysis of randomised clinical trials. Lancet, 2007;370:11291136.
  • 11
    Nissen SE, Wolski K. Rosiglitazone revisited: an updated meta-analysis of risk for myocardial infarction and cardiovascular mortality. Arch Intern Med, 2010;170:11911201.
  • 12
    Monami M, Cremasco F, Lamanna C et al. Glucagon-like peptide-1 receptor agonists and cardiovascular events: a meta-analysis of randomized controlled trials. Exp Diabetes Res, 2011;2011:215764.
  • 13
    Best JH, Hoogwerf BJ, Herman WH et al. Risk of cardiovascular disease events in patients with type 2 diabetes prescribed the glucagon-like peptide 1 (GLP-1) receptor agonist exenatide twice daily or other glucose-lowering therapies: a retrospective analysis of the LifeLink database. Diabetes Care, 2011;34:9095.
  • 14
    Thorlund K, Devereaux PJ, Wetterslev J et al. Can trial sequential monitoring boundaries reduce spurious inferences from meta-analyses? Int J Epidemiol, 2009;38:276286.
  • 15
    Wetterslev J, Thorlund K, Brok J et al. Trial sequential analysis may establish when firm evidence is reached in cumulative meta-analysis. J Clin Epidemiol, 2008;61:6475.
  • 16
    Brok J, Thorlund K, Gluud C et al. Trial sequential analysis reveals insufficient information size and potentially false positive results in many meta-analyses. J Clin Epidemiol, 2008;61:763769.
  • 17
    Thorlund K, Engstrom J, Wetterslev J et al. User manual for trial sequential analysis (TSA). Copenhagen, Denmark: Copenhagen Trial Unit, Centre for Clinical Intervention Research, 2011. p1115. www.ctu.dk/tsa
  • 18
    Bangalore S, Kumar S, Kjeldsen SE et al. Antihypertensive drugs and risk of cancer: network meta-analyses and trial sequential analyses of 324,168 participants from randomised trials. Lancet Oncol, 2011;12:6582.
  • 19
    Hemmingsen B, Lund SS, Gluud C et al. Intensive glycaemic control for patients with type 2 diabetes: systematic review with meta-analysis and trial sequential analysis of randomised clinical trials. BMJ, 2011;343:d6898.
  • 20
    Novo Nordisk. Liraglutide (injection) for the Treatment of Patients with Type 2 Diabetes: NDA22-341. Endocrine and Metabolic Drug Advisory Committee 2 April 2009. Available at: http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/ Drugs/EndocrinologicandMetabolicDrugsAdvisoryCommittee/UCM148659.pdf (accessed 1 December 2011).
  • 21
    Jadad AR, Moore RA, Carroll D et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials, 1996;17:112.
  • 22
    Keus F, Wetterslev J, Gluud C et al. Robustness assessments are needed to reduce bias in meta-analyses that include zero-event randomized trials. Am J Gastroenterol, 2009;104:546551.
  • 23
    Rucker G, Schwarzer G, Carpenter J et al. Why add anything to nothing? The arcsine difference as a measure of treatment effect in meta-analysis with zero cells. Stat Med, 2009;28:721738.
  • 24
    Friedrich JO, Adhikari NK, Beyene J. Inclusion of zero total event trials in meta-analyses maintains analytic consistency and incorporates all available data. BMC Med Res Methodol, 2007;7:5.
  • 25
    Ratner R, Han J, Nicewarner D et al. Cardiovascular safety of exenatide BID: an integrated analysis from controlled clinical trials in participants with type 2 diabetes. Cardiovasc Diabetol, 2011;10:22.
  • 26
    Shyangdan DS, Royle P, Clar C et al. Glucagon-like peptide analogues for type 2 diabetes mellitus. Cochrane Database Syst Rev, 2011; 10: CD006423.
  • 27
    Monami M, Marchionni N, Mannucci E. Glucagon-like peptide-1 receptor in type 2 diabetes: a meta-analysis of randomized controlled trials. Eur J Endocrinol, 2009;160:909917.
  • 28
    Nauck MA, Duran S, Kim D et al. A comparison of twice-daily exenatide and biphasic insulin aspart in patients with type 2 diabetes who were suboptimally controlled with sulfonylurea and metformin: a non-inferiority study. Diabetologia, 2007;50:259267.
  • 29
    Li WX, Guo JF, Tian JH et al. Glucagon-like peptide-1 receptor agonists versus insulin glargine for type 2 diabetes mellitus: a systematic review and meta-analysis of randomized controlled trials. Curr Ther Res Clin Exp, 2010;71:211238.
  • 30
    Wang Y, Li L, Yang M et al. Glucagon-like peptide-1 receptor agonists versus insulin in inadequately controlled patients with type 2 diabetes mellitus: a meta-analysis of clinical trials. Diabetes Obes Metab, 2011;13:972981.
  • 31
    Bolen S, Feldman L, Vassy J et al. Systematic review: comparative effectiveness and safety of oral medications for type 2 diabetes mellitus. Ann Intern Med, 2007;147:386399.
  • 32
    Arakawa M, Mita T, Azuma K et al. Inhibition of monocyte adhesion to endothelial cells and attenuation of atherosclerotic lesion by a glucagon-like peptide-1 receptor agonist, exendin-4. Diabetes, 2010;59:10301037.
  • 33
    Koska J, Schwartz EA, Mullin MP et al. Improvement of postprandial endothelial function after a single dose of exenatide in individuals with impaired glucose tolerance and recent-onset type 2 diabetes. Diabetes Care, 2010;33:10281030.
  • 34
    Timmers L, Henriques JP, de Kleijin DP et al. Exenatide reduces infarct and improves cardiac function in a porcine model of ischemia and reperfusion injury. J Am Coll Cardiol, 2009;53:501510.
  • 35
    Kristensen J, Mortensen UM, Schmidt M et al. Lack of cardioprotection from subcutaneously and preischemic administered liraglutide in a closed chest porcine ischemia reperfusion model. BMC Cardiovasc Disord, 2009;23:31.
  • 36
    Liu Q, Anderson C, Broyde A et al. Glucagon-like peptide-1 and the exenatide analogue AC3174 improve cardiac function, cardiac remodeling, and survival in rats with chronic heart failure. Cardiovasc Diabetol, 2010;9:76.
  • 37
    Johansen OE, Neubacher D, von Eynatten M et al. Cardiovascular safety with linagliptin in patients with type 2 diabetes mellitus: a pre-specified, prospective, and adjudicated meta-analysis of a phase 3 programme. Cardiovasc Diabetol, 2012;11:3.
  • 38
    Alderson P. Absence of evidence is not evidence of absence. BMJ, 2004;328:476477.
  • 39
    ClinicalTrials.gov, a service of the U.S. National Institutes of Health. http://www.clinicaltrial.gov.

Supporting Information

  1. Top of page
  2. Summary
  3. What is known and objective
  4. Methods
  5. Results
  6. Discussion
  7. What is new and conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
jcpt12102-sup-0001-TableS1-S2.docWord document249K

Table S1 Characteristics of the studies included in the meta-analysis

Table S2 Outcome variables of the studies included in the meta-analysis

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.