Author + information
- David Fitchett, MD∗ (, )
- Silvio E. Inzucchi, MD,
- John M. Lachin, ScD,
- Christoph Wanner, MD,
- Philippe van de Borne, MD,
- Michaela Mattheus, Dipl Biomath,
- Odd Erik Johansen, MD, PhD,
- Hans J. Woerle, MD,
- Uli C. Broedl, MD,
- Jyothis T. George, MBBS, PhD,
- Bernard Zinman, MD,
- on behalf of the EMPA-REG OUTCOME Investigators
- ↵∗Division of Cardiology, St. Michael’s Hospital, University of Toronto, 30 Bond Street, Suite 7037, Toronto, Ontario M5B 1W8, Canada
Diabetes is associated with a significant decrease in life expectancy (1). In the EMPA-REG OUTCOME trial (NCT01131676), empagliflozin given in addition to standard of care reduced the risk of cardiovascular death by 38% (hazard ratio [HR]: 0.62; 95% confidence interval [CI]: 0.49 to 0.77; p < 0.0001) and all-cause mortality by 32% (HR: 0.68; 95% CI: 0.57 to 0.82; p <0.0001) versus placebo in patients with type 2 diabetes and established cardiovascular disease (2). These results led to the first approval of a cardiovascular death prevention indication for a glucose-lowering drug in the United States. We now report the time course, subgroup, and sensitivity analyses to interrogate the robustness of our findings.
Clinical event committees prospectively categorized deaths as cardiovascular or noncardiovascular according to pre-specified criteria. Cardiovascular deaths were classified into fatal myocardial infarction, fatal stroke, death due to heart failure (worsening of heart failure or cardiogenic shock), sudden death, or other cardiovascular cause. In this population with established cardiovascular disease, when there were insufficient data for the adjudication committees to attribute a cause of death according to pre-specified criteria, deaths were categorized as “presumed cardiovascular death,” as has been the convention in large long-term outcome trials involving patients who are at high cardiovascular risk. Cox proportional hazards models compared the pooled empagliflozin and placebo groups using the HR in a modified intent-to-treat cohort of patients treated with ≥1 dose of study drug. Models included treatment, age, sex, baseline body mass index, baseline hemoglobin A1c, baseline estimated glomerular filtration rate, and region as factors. We present HRs and 95% CIs derived at each time point following randomization until the last observation of the last patient. All events until the respective day were considered, and patients without events were censored at the respective day or the last day they were known to be free of the outcome (i.e., when they were last known to be alive or had died due to a noncardiovascular cause), whichever was earlier. Post hoc subgroup analyses were conducted by type of cardiovascular disease at baseline and included additional effects for a subgroup factor and a treatment by subgroup factor interaction.
Separation of the cumulative incidence curves for cardiovascular death with empagliflozin versus placebo appeared early and persisted for the duration of the trial (Figure 1A). HRs stabilized as the number of deaths increased over time. The reduction in risk of cardiovascular death was consistent across baseline characteristics (2), including type of cardiovascular disease (Figure 1B). All categories of cardiovascular death contributed to the risk reduction with empagliflozin, with the most frequent modes being sudden death (29.4%), death due to heart failure (11.7%), and presumed cardiovascular death (40.1%). When presumed cardiovascular death was excluded, the reduction in cardiovascular death with empagliflozin versus placebo remained significant (HR: 0.59; 95% CI: 0.44 to 0.79; p < 0.001) (Figure 1B). The HR for presumed cardiovascular death was 0.66 (95% CI: 0.46 to 0.94; p = 0.0218).
Although cardiovascular death was not explicitly part of the confirmatory testing hierarchy in the EMPA-REG OUTCOME trial, a favorable result on a major outcome such as mortality is difficult to ignore, as highlighted in recent U.S. Food and Drug Administration guidance on multiple endpoints in clinical trials (3). We consider this reduction in cardiovascular death in the EMPA-REG OUTCOME trial to be robust due to consistency across doses, subgroups, and analysis populations, and a clinically relevant effect size. In addition, all categories of cardiovascular death showed a favorable point estimate and remained significant in a sensitivity analysis excluding presumed cardiovascular death. Bayesian models have shown strong evidence that empagliflozin reduces mortality (4). These observations are relevant in the context of clinical care and the design of clinical trials.
The biological underpinnings of the reduction in cardiovascular death with empagliflozin remain unknown. The early diversion in the event curves suggests that the predominant mechanism may relate to its hemodynamic effects. It is therefore tempting to speculate that the reduction in volume and sodium load, leading to reduced ventricular filling pressure and cardiac workload, could be an important mechanism (5). Translational studies of the physiological effects of empagliflozin and ongoing clinical trials in patients with heart failure should help inform our understanding of the mechanisms underlying the reduction in cardiovascular death with empagliflozin.
Please note: This trial was funded by the Boehringer Ingelheim & Eli Lilly and Company Diabetes Alliance. Dr. Fitchett has received steering committee honoraria from Boehringer Ingelheim; and has received CME and consulting honoraria from Sanofi, Merck, Amgen, AstraZeneca, Eli Lilly and Company, and Boehringer Ingelheim. Dr. Inzucchi has received honoraria from AstraZeneca, Boehringer Ingelheim, Daiichi-Sankyo, Eisai, Janssen, Intarcia Therapeutics, Inc., Novo Nordisk, Sanofi/Lexicon, and vTv Therapeutics. Dr. Lachin has received honoraria from AbbVie, AstraZeneca, Boehringer Ingelheim, Janssen, and Merck; and has served on the publications committees of the TECOS study (funded by Merck) and the EMPA-REG OUTCOME Study (funded by Boehringer Ingelheim). Dr. Wanner has received honoraria from Boehringer Ingelheim and Janssen; and has served as an advisory board member for Boehringer Ingelheim. Dr. van de Borne has received a research grant from AstraZeneca and Medtronic; and has received honoraria from Boehringer Ingelheim, Novo Nordisk, AstraZeneca, Amgen, and Sanofi. Drs. Mattheus, Johansen, Woerle, Broedl, and George are employed by Boehringer Ingelheim. Dr. Zinman has received research grants awarded to his institution from Boehringer Ingelheim, AstraZeneca, and Novo Nordisk; and has received honoraria from Janssen, Sanofi, Eli Lilly and Company, Boehringer Ingelheim, Novo Nordisk, and Merck. Editorial assistance, supported financially by Boehringer Ingelheim, was provided by Elizabeth Ng and Wendy Morris of FleishmanHillard Fishburn, London, UK, during the preparation of this paper.
- 2018 American College of Cardiology Foundation
- ↵U.S. Department of Health and Human Services Food and Drug Administration. Multiple endpoints in clinical trials guidance for industry. 2017. Available at: www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm536750.pdf. Accessed November 22, 2017.
- Kaul S.
- Sattar N.,
- McLaren J.,
- Kristensen S.L.,
- Preiss D.,
- McMurray J.J.