DPP4i initiators were on average younger (60 vs 63 years), used metformin monotherapy for longer (2.7 vs 2.3 years), had a higher BMI (34 vs 32 kg/m2), and had a lower A1c (8.5% vs 8.8%), compared to SU initiators; similar differences were seen with GLP-1RA initiators (Table 1). A total of 21,848 (57%) patients were linked to hospitalization (HES) and death certificate (ONS) data of which 15% initiated DPP4i, 1% GLP-1RA, 69% SU, 12% TZD, 2% insulin, and 1% OTH (Supplemental Figure 1 and Supplemental Table 1, http://links.lww.com/MD/B69).
3.1 All-cause mortality
There were a total of 2684 deaths in the study period, 1500 of those events occurred while patients were exposed to their 2nd antidiabetic agent following metformin monotherapy. Mortality rates were 8.2 deaths per 1000 person-years (95% confidence interval [CI] 6.6–10.1) for DPP4i initiators, 19.1 deaths per 1000 person-years (95% CI 18.0–20.2) for SU initiators, 7.2 deaths per 1000 person-years (95% CI 5.8–8.9) for TZD initiators, 70.9 deaths per 1000 person-years (95% CI 59.1–85.0) for insulin initiators, and 18.7 deaths per 1000 person-years (95% CI 11.4–30.5) for OTH (Table 2). After adjusting for potential confounders, among patients who started metformin monotherapy, the initiation of a DPP4i compared to an SU was associated with a significant 42% reduction in mortality (absolute rate difference = 11 events per 1000 person-years, aHR = 0.58, 95% CI 0.46–0.73, P < 0.001) (Table 2).
3.2 Major adverse cardiovascular events
Over the study period, there were a total of 1854 MACE that occurred within the HES and ONS linked population (n = 21,848), 1455 of those events occurred in patients that initiated a 2nd-line antidiabetic agent. For DPP4i initiators, there were 19.1 MACE per 1000 person-years (95% CI 15.7–23.3); for GLP-1RA initiators, 15.9 MACE per 1000 person-years (95% CI 7.6–33.4); for SU initiators, 33.1 MACE per 1000 person-years (95% CI 31.2–35.1); for TZD initiators, 20.7 MACE per 1000 person-years (95% CI 17.5–24.5); for insulin initiators, 63.3 MACE per 1000 person-years (95% CI 48.7–82.3); and for initiators of OTH, 28.4 MACE per 1000 person-years (95% CI 16.8–47.9) (Table 3). Compared to 2nd-line SU use, 2nd-line DPP4i use was associated with a statistically significant reduction in MACE (absolute risk difference = 14 events per 1000 person-years, aHR = 0.64, 95% CI 0.52–0.80, P < 0.001). GLP1RAs were not associated with a reduction in MACE (absolute risk difference = 17 events per 1000 person-years, aHR = 0.73, 95% CI 0.34–1.55, P = 0.44).
3.3 Secondary outcomes
Initiators of a DPP4i as a 2nd-line agent compared to SU were associated with a statistically significant reduction in risk for 5 of the 6 secondary outcomes: unstable angina (aHR = 0.64 95% CI 0.52–0.80), arrhythmia (aHR = 0.66 95% CI 0.55–0.78), heart failure (aHR = 0.57 95% CI 0.42–0.75), myocardial infarction (aHR = 0.66 95% CI 0.47–0.94), and urgent revascularization (aHR = 0.58 95% CI 0.52–0.64), but not stroke (aHR = 0.97 95% CI 0.80–1.17). There was also a significant reduction in risk for cardiovascular death (absolute rate difference = 6 events per 1000 person-years, aHR = 0.16, 95% CI 0.06–0.44). There were no significant subgroup treatment effects observed for a priori defined patient characteristics hypothesized to have a potential impact on the risk of mortality (P-value > 0.15 for all interaction terms) (Figure 2 and Supplemental Figure 2, http://links.lww.com/MD/B69).
3.4 Sensitivity analysis
All sensitivity analyses supported the robustness of our primary findings (Figure 3). A significant reduction in mortality was also found when 2nd-line DPP4i users were compared to 2nd-line insulin users (aHR = 0.21, 95% CI 0.16–0.28, P < 0.001) and with 2nd-line TZD (aHR = 0.74, 95% CI 0.54–0.99, P = 0.04). Anytime use of a DPP4i after metformin monotherapy was also associated with a reduction in mortality compared to SU use (aHR = 0.56, 95% CI 0.45–0.69). Anytime use of a GLP1RA was not associated with a difference in mortality (aHR = 0.70, 95% CI 0.38–1.27). Statistically significant differences in MACE were observed for 2nd-line DPP4i compared to 2nd-line insulin (aHR = 0.43, 95% CI 0.30–0.60, P < 0.001) and 2nd-line TZD (aHR = 0.68, 95% CI 0.52–0.89, P = 0.005) users. Accounting for anytime use of a DPP4i after metformin monotherapy compared to anytime use of an SU, a significant reduction in the risk of MACE was observed (aHR = 0.67, 95% CI 0.55–0.82, P < 0.001). Anytime exposure of a GLP1RA was not associated with a reduction in MACE (aHR = 0.75, 95% CI 0.46–1.24).
4.1 Main results
Using a large population-based cohort of new-users of metformin monotherapy, we found that initiation of 2nd-line DPP4i was associated with a 42% reduction in mortality and 36% reduction in the composite of myocardial infarction, stroke, or cardiovascular death compared to 2nd-line initiation of SU. Similarly, 2nd-line DPP4i use was associated with a reduction in death and cardiovascular events compared to 2nd-line insulin and TZD use. With the exception of stroke (no association), DPP4is were also associated with a decreased risk of each individual cardiovascular endpoint examined. Second-line initiation of GLP-1RAs was not associated with an increased or decreased risk of MACE compared to SU therapy.
These findings are consistent with the putative cardioprotection reported in preclinical and clinical studies of incretins.[5,23,24] Alternatively, our findings could be a result of adverse cardiovascular effects of the SU comparator antidiabetic therapies we studied. The cardiotoxicity of SUs has been an area of considerable debate for over 40 years and is an area of continued investigation. Mechanisms by which SUs may affect cardiovascular risk include attenuation of ischemic preconditioning, hyperinsulinemia, weight gain, and hypoglycemia. However, in sensitivity analyses where the comparator groups were insulin or TZDs, the findings were similar suggesting to us that cardioprotection may be the more likely explanation.
4.2 Comparison with other literature
Recent randomized controlled trials have found no statistical or clinical differences between DPP4is and placebo for the incidence of major cardiovascular outcomes including myocardial infarction, stroke, or cardiovascular death.[6–8] Likewise, a recent cardiovascular outcome trial, ELIXA, found no differences in the rate of major adverse cardiovascular events for the GLP-1RA lixisenatide compared to placebo.
Previous observational studies evaluating DPP4is have been inconsistent in their findings.[10–16,27] For example, using a large US-based claims and integrated laboratory database, Eurich et al conducted a cohort study including 72,738 new-users of oral antidiabetic drugs of which 8032 used sitagliptin. They found no difference in the risk of their primary composite outcome of all-cause mortality or hospitalization for those exposed to sitagliptin versus other antidiabetic drugs (aHR = 0.98, 95% CI 0.91–1.06). There were 32 deaths in the sitagliptin group thereby limiting the power to detect relative differences in the rate of death compared to nonsitagliptin users. A time-varying approach was used to categorize antidiabetic drug exposure over the entire observation period that assumes exchangeability of exposure categories irrespective of ordering or gaps in antidiabetic exposure. Conversely, Rathman et al used the German Disease Analyzer database (IMS HEALTH) and found new use of DPP4i was associated with a decreased risk of macrovascular events (aHR = 0.74, 95% CI 0.67–0.82), defined as primary care diagnoses for coronary heart disease, myocardial infarction, stroke, and peripheral vascular disease; and Ou et al used administrative data from Taiwan and found DPP4is were associated with decreased mortality (aHR = 0.63, 95% CI 0.55–0.72), MACE (aHR = 0.63, 95% CI 0.55–0.83), and ischemic stroke (aHR = 0.43, 95% CI 0.51–0.81) but no difference in the risk of myocardial infarction.
In fact, there have been 3 prior studies using the CPRD database to examine this important clinical question (using different analytic methods and smaller cohorts than ours), reporting inconsistent effects of DPP4i.[12–14] The 1st study included 27,457 new metformin monotherapy users with type 2 diabetes who initiated a 2nd antidiabetic therapy. At least 180 days of metformin exposure was an inclusion criterion, creating a period of immortal time. They found reported combination MET and DPP4i (n = 1455) therapy was not associated with all-cause mortality (aHR = 0.61, 95% CI 0.29–1.30) or a composite of a myocardial infarction or stroke (aHR = 1.02, 95% CI 0.50–2.12). There were, however, only 9 deaths, 4 myocardial infarctions, and 4 strokes, in the metformin and DPP4i group versus 86 deaths and 98 MACE events in our study. The 2nd CPRD study used different inclusion criteria and exposure definitions and found a statistically significant decrease in mortality (aHR = 0.74, 95% CI 0.58–0.92) and nonsignificant decrease in MACE (aHR = 0.76, 95% CI 0.47–1.20) for MET-DPP4i users compared to MET-SU users. The 3rd study, by Yu et al, found the use of a DPP4i with metformin was associated with a reduction in the risk of a composite of all-cause mortality, myocardial infarction, or stroke (aHR = 0.62, 95% CI 0.40–0.98) compared to combination use of metformin an SU. However, this study was also limited in the number of events (12 MACE and 13 deaths in the Metformin-DPP4i group) and did not include cardiovascular-related death within their MACE composite.
Similarly, observational studies have shown mixed results regarding the association between GLP-1RA use and cardiovascular events.[28–31] Although we had limited statistical power to evaluate the association between GLP-1RAs and cardiovascular events, our results are consistent with findings from other studies which found GLP-1RAs were not associated with an increased or decreased risk of cardiovascular events compared to SUs. Patorno et al used a large US commercial database to compare the incidence of a composite cardiovascular endpoint (hospitalization for acute MI, unstable angina, stroke, and coronary revascularization) between GLP-1RA initiators and SU initiators and found no differences in risk (aHR = 0.86, 95% CI 0.69–1.08). Using a Danish database, Mogensen et al also found no differences in cardiovascular risk between GLP-1RAs and SUs (aHR = 0.82, 95% CI 0.55–1.21). However, others have reported cardioprotective effects of GLP-1RAs, most likely due to differences in comparator groups and other methodological differences. For example, Paul et al used insulin as a comparator group, and found the GLP1RA, exenatide, was associated with a decreased risk of heart failure, MI, and stroke. Best et al also found exenatide was associated with a decreased risk of cardiovascular disease, however they used a comparator group with mixture of antidiabetic agents. All of the aforementioned studies lacked sufficient power to evaluate individual cardiovascular endpoints with precision.
How does our study add to this evidence? First, our design precludes immortal time bias. Second, we had many more events and a larger population with longer follow-up time. Third, our study includes a large number of outcomes for all-cause mortality and major adverse cardiovascular outcome events in DPP4i users, thereby providing sufficient power to minimize the risk of a type 2 statistical error. Importantly, our definition of MACE includes cardiovascular related mortality using death certificate data through linkage with ONS data. Finally, our study used the most clinically relevant study population – guideline concordant new metformin users who then initiated 1 of 6 widely available 2nd-line antidiabetic agents.
Although our cohort study used rigorous design and analytic methods, several limitations must be acknowledged. First, as in all observational studies we cannot rule out the potential for residual confounding, or unknown or unmeasured confounders (e.g., ethnicity, family history, and activity level) as alternate explanation for our findings; however, we adjusted for numerous potential confounders (i.e., A1c, smoking, deprivation index, BMI, SBP, eGFR, etc.) that are often not available within computerized administrative databases. Second, confounding by indication is always a concern when evaluating intended outcomes of drug therapies using observational study designs. We used several design and analysis techniques to minimize this bias including restriction of subject to metformin monotherapy users, a new user design, an active comparator, and rigorous methods to control for confounding. Similarly, channeling may occur whereby patients differ in certain characteristics (e.g., comorbidities) that may affect decision to prescribe on drug versus another. Certainly, an individual's weight, financial status, comorbidities, and kidney function among other characteristics would be expected to influence whether an incretin-based therapy, insulin, TZD, or an SU was prescribed. Third, we assume that drug prescription is a surrogate marker for consumption. We recognize this method may overestimate actual exposure since primary and secondary adherence failure cannot be assessed and this may bias our observation towards the null hypothesis.
Our findings suggest that the guideline-concordant initiation of 2nd-line DPP4i following metformin monotherapy is associated with lower risk of all-cause mortality and major cardiovascular events compared to 2nd-line SU. It is noteworthy that almost 70% of the population received an SU as add-on therapy in our cohort. Until adequately powered active comparator trials of 2nd-line antidiabetic agents in patients on metformin monotherapy are reported, the evidence to date and now including our study suggests that DPP4is are the safest of the 2nd-line agents to use following failure of metformin monotherapy. Whether our findings are a result of a protective effect of DPP4i or a harmful effect of SUs cannot be determined by this study design, but the relative mortality reductions seen compared to other antidiabetic agents tends to favor the former.
This study is based in part on data from the Clinical Practice Research Datalink obtained under licence from the UK Medicines and Healthcare products Regulatory Agency. However, the interpretation and conclusions contained in this study are those of the author/s alone.
1. Canadian Diabetes Association Clinical Expert, Committee. Canadian Diabetes Association 2013 Clinical Practice Guidelines for the Prevention and Management of Diabetes in Canada. Can J Diabetes
2013; 37 (suppl 1):1–227.
2. Canadian Agency for Drugs and Technologies in Health. Current Utilization of Second and Third-Line Therapies in Patients with Type 2 Diabetes. Ottawa:2010.
3. Inzucchi SE, Bergenstal RM, Buse JB, et al Management of hyperglycemia in type 2 diabetes: a patient-centered approach position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care
4. Gamble JM, Clarke A, Myers KJ, et al Incretin-based medications for type 2 diabetes: an overview of reviews. Diabetes Obes Metab
5. Okerson T, Chilton RJ. The cardiovascular effects of GLP-1 receptor agonists. Cardiovasc Ther
6. White WB, Cannon CP, Heller SR, et al Alogliptin after acute coronary syndrome in patients with type 2 diabetes. N Engl J Med
7. Scirica BM, Bhatt DL, Braunwald E, et al Saxagliptin, cardiovascular outcomes in patients with type 2 diabetes mellitus
. N Engl J Med
8. Green JB, Bethel MA, Armstrong PW, et al Effect of sitagliptin on cardiovascular outcomes in type 2 diabetes. N Engl J Med
9. Zannad F, Cannon CP, Cushman WC, et al Heart failure and mortality
outcomes in patients with type 2 diabetes taking alogliptin versus placebo in EXAMINE: a multicentre, randomised, double-blind trial. Lancet
10. Eurich DT, Simpson S, Senthilselvan A, et al Comparative safety and effectiveness of sitagliptin in patients with type 2 diabetes: retrospective population based cohort study. BMJ
11. Rathmann W, Kostev K, Gruenberger JB, et al Treatment persistence, hypoglycaemia and clinical outcomes in type 2 diabetes patients with dipeptidyl peptidase-4 inhibitors and sulphonylureas: a primary care database analysis. Diabetes Obes Metab
12. Morgan CL, Poole CD, Evans M, et al What next after metformin? A retrospective evaluation of the outcome of second-line, glucose-lowering therapies in people with type 2 diabetes. J Clin Endocrinol Metab
13. Morgan CL, Mukherjee J, Jenkins-Jones S, et al Combination therapy with metformin plus sulphonylureas versus metformin plus DPP-4 inhibitors: association with major adverse cardiovascular events and all-cause mortality
. Diabetes Obes Metab
14. Yu OHY, Yin H, Azoulay L. The combination of dpp-4 inhibitors versus sulfonylureas with metformin after failure of first-line treatment in the risk for major cardiovascular events and death. Can J Diabetes
15. Weir DL, McAlister FA, Senthilselvan A, et al Sitagliptin use in patients with diabetes and heart failure. JACC Heart Fail
16. Ou S-M, Shih C-J, Chao P-W, et al Effects on clinical outcomes of adding dipeptidyl peptidase-4 inhibitors versus sulfonylureas to metformin therapy in patients with type 2 diabetes mellitus
. Ann Intern Med
17. Rodriguez LA, Gutthann SP. Use of the UK general practice research database for pharmacoepidemiology. Br J Clin Pharmacol
18. Herrett E, Thomas SL, Schoonen WM, et al Validation and validity of diagnoses in the General Practice Research Database: a systematic review. Br J Clin Pharmacol
19. Jick SS, Kaye JA, Vasilakis-Scaramozza C, et al Validity of the general practice research database. Pharmacotherapy
20. Williams T, Van Staa T, Puri S, et al Recent advances in the utility and use of the General Practice Research Database as an example of a UK Primary Care Data resource. Ther Adv Drug Saf
21. Khan NF, Harrison SE, Rose PW. Validity of diagnostic coding within the General Practice Research Database: a systematic review. Br J Gen Pract
22. Pöge U, Gerhardt T, Palmedo H, et al MDRD equations for estimation of GFR in renal transplant recipients. Am J Transplant
23. Gallwitz B. Preclinical and clinical data on extraglycemic effects of GLP-1 receptor agonists. Rev Diabet Stud
24. Inzucchi SE, McGuire DK. New drugs for the treatment of diabetes: part II: incretin-based therapy and beyond. Circulation
25. Abdelmoneim AS, Eurich DT, Light PE, et al Cardiovascular safety of sulphonylureas: over 40 years of continuous controversy without an answer. Diabetes Obes Metab
26. Pfeffer MA, Claggett B, Diaz R, et al Lixisenatide in patients with type 2 diabetes and acute coronary syndrome. N Engl J Med
27. Yu OHY, Filion KB, Azoulay L, et al Incretin-based drugs and the risk of congestive heart failure. Diabetes Care
28. Patorno E, Everett B, Goldfine A, et al Comparative cardiovascular safety of glucagon-like peptide-1 receptor agonists versus other antidiabetic drugs
in routine care: a cohort study. Diabetes Obes Metab
2016; online early May 2, 2016. doi: 10.1111/dom.12665.
29. Mogensen UM, Andersson C, Fosb⊘l EL, et al Cardiovascular safety of combination therapies with incretin-based drugs and metformin compared with a combination of metformin and sulphonylurea in type 2 diabetes mellitus
–a retrospective nationwide study. Diabetes Obes Metab
30. 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
31. Paul SK, Klein K, Maggs D, et al The association of the treatment with glucagon-like peptide-1 receptor agonist exenatide or insulin with cardiovascular outcomes in patients with type 2 diabetes: a retrospective observational study. Cardiovasc Diabetol
32. Suissa S. Effectiveness of inhaled corticosteroids in chronic obstructive pulmonary disease: immortal time bias in observational studies. Am J Respir Crit Care Med
antidiabetic drugs; cardiovascular diseases; cohort studies; dipeptidyl-peptidase IV inhibitors; mortality; sulfonylurea compounds; type 2 diabetes mellitus
Supplemental Digital Content
Copyright © 2016 The Authors. Published by Wolters Kluwer Health, Inc. All rights reserved.