Heart Failure. Heart Failure, Saxagliptin, and Diabetes Mellitus: Observations from the SAVOR-TIMI 53 Randomized Trial

Transcription

1 Heart Failure Heart Failure, Saxagliptin, and Diabetes Mellitus: Observations from the SAVOR-TIMI 53 Randomized Trial Benjamin M. Scirica, MD, MPH; Eugene Braunwald, MD; Itamar Raz, MD; Matthew A. Cavender, MD, MPH; David A. Morrow, MD, MPH; Petr Jarolim, MD, PhD; Jacob A. Udell, MD, MPH; Ofri Mosenzon, MD; KyungAh Im, PhD; Amarachi A. Umez-Eronini, MPH; Pia S. Pollack, MD; Boaz Hirshberg, MD; Robert Frederich, MD, PhD; Basil S. Lewis, MD; Darren K. McGuire, MD, MHSc; Jaime Davidson, MD; Ph. Gabriel Steg, MD; Deepak L. Bhatt, MD, MPH; for the SAVOR-TIMI 53 Steering Committee and Investigators* Background Diabetes mellitus and heart failure frequently coexist. However, few diabetes mellitus trials have prospectively evaluated and adjudicated heart failure as an end point. Methods and Results A total of patients with type 2 diabetes mellitus and a history of, or at risk of, cardiovascular events were randomized to saxagliptin or placebo (mean follow-up, 2.1 years). The primary end point was the composite of cardiovascular death, myocardial infarction, or ischemic stroke. Hospitalization for heart failure was a predefined component of the secondary end point. Baseline N-terminal pro B-type natriuretic peptide was measured in patients. More patients treated with saxagliptin (289, 3.5%) were hospitalized for heart failure compared with placebo (228, 2.8%; hazard ratio, 1.27; 95% confidence intercal, ; P=0.007). Corresponding rates at 12 months were 1.9% versus 1.3% (hazard ratio, 1.46; 95% confidence interval, ; P=0.002), with no significant difference thereafter (time-varying interaction, P=0.017). Subjects at greatest risk of hospitalization for heart failure had previous heart failure, an estimated glomerular filtration rate 60 ml/min, or elevated baseline levels of N-terminal pro B-type natriuretic peptide. There was no evidence of heterogeneity between N-terminal pro B-type natriuretic peptide and saxagliptin (P for interaction=0.46), although the absolute risk excess for heart failure with saxagliptin was greatest in the highest N-terminal pro B-type natriuretic peptide quartile (2.1%). Even in patients at high risk of hospitalization for heart failure, the risk of the primary and secondary end points were similar between treatment groups. Conclusions In the context of balanced primary and secondary end points, saxagliptin treatment was associated with an increased risk or hospitalization for heart failure. This increase in risk was highest among patients with elevated levels of natriuretic peptides, previous heart failure, or chronic kidney disease. Clinical Trial Registration URL: Unique identifier: NCT (Circulation. 2014;130: ) Diabetes mellitus (DM) adds incremental risk of the development and subsequent exacerbation of heart failure even after adjusting for common risk factors, such as ischemic Key Words: diabetes mellitus heart failure saxagliptin heart disease and hypertension. 1 Although incompletely understood, cardiac metabolic dysregulation of glycolysis and fatty acid oxidation observed in the heart of patients with Continuing medical education (CME) credit is available for this article. Go to to take the quiz. Received April 4, 2014; accepted August 20, From the Thrombolysis in Myocardial Infarction (TIMI) Study Group, Cardiovascular Division (B.M.S., E.B., M.A.C., D.A.M., K.I., A.A.U.-E., D.L.B.) and Department of Pathology (P.J.), Brigham and Women s Hospital and Harvard Medical School, Boston, MA; Diabetes Unit (I.R., O.M.), Division of Internal Medicine, Hadassah Hebrew University Hospital, Jerusalem, Israel; Cardiovascular Division (J.A.U.), Women s College Hospital and Toronto General Hospital, University of Toronto, Toronto, Canada; AstraZeneca Research and Development (P.S.P., B.H.), Gaithersburg, MD; Bristol-Myers Squibb (R.F.), Princeton, NJ; Cardiovascular Research Institute (B.S.L.), Lady Davis Carmel Medical Center and Ruth and Bruce Rappaport School of Medicine, Technion Israel Institute of Technology, Haifa, Israel; Cardiovascular Medicine (D.K.M.) and Division of Endocrinology, Department of Internal Medicine (J.D.), University of Texas Southwestern Medical Center, Dallas, TX; University Hospital Department (G.S.), Département Hospitalo-Universitaire FIRE (Fibrosis, Inflammation, Remodelling), Institut National de la Santé et de la Recherche Médicale (INSERM) Unit 1148, Université Paris-Diderot, and Hôpital Bichat, Assistance Publique- Hôpitaux de Paris, Paris, France (P.G.S.), Imperial College, Institute of Cardiovascular Medicine and Science, Royal Brompton Hospital, London, UK. Guest Editor for this article was Michael E. Farkouh, MD. *A complete list of the SAVOR-TIMI 53 Steering Committee and Investigators can be found in the Online Data Supplement. The online-only Data Supplement is available with this article at /-/DC1. Correspondence to Benjamin M. Scirica, MD, MPH, TIMI Study Group, Cardiovascular Division, Brigham and Women s Hospital, 75 Francis St, Boston, MA bscirica@partners.org 2014 American Heart Association, Inc. Circulation is available at DOI: /CIRCULATIONAHA

2 1580 Circulation October 28, 2014 DM probably impairs cardiac function and causes additional myocardial damage. 2 5 In addition, accelerated coronary atherosclerosis is likely to play a role. 6 Moreover, the choice of antihyperglycemic agents in those patients with DM with heart failure or at risk of developing it remains challenging. Some agents, such as thiazolidinediones and dual peroxisome proliferator-activated receptor α/γ agonists, increase plasma volume and exacerbate heart failure, 7 11 whereas sulfonylureas 12 and insulin potentially exacerbate the dysregulation of myocardial metabolism and worsen left ventricular function. 13,14 Metformin, once felt to be contraindicated in patients with heart failure, is now considered to be 1 of the safest options, despite the absence of large randomized comparisons. 15,16 Despite the observational relationship between glycemic control and the risk of heart failure, there is no evidence that improved glycemic control modifies this risk. 17 Therefore, identification of antihyperglycemic agents that can be used safely in patients with heart failure or at risk of developing heart failure remains an important unmet clinical need. Editorial see p 1565 Clinical Perspective on p 1588 Saxagliptin is a selective dipeptidyl peptidase-4 (DPP-4) inhibitor. 18 The Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus Thrombolysis in Myocardial Infarction 53 (SAVOR-TIMI 53) trial was designed to evaluate the long-term cardiovascular efficacy and safety of saxagliptin in patients with DM at risk of cardiovascular events. 19 Over a median of 2.1 years follow-up, saxagliptin neither increased nor decreased the risk of the primary or secondary composite end points; however, there was an unexpected 27% increased relative risk (and 0.7% absolute risk over 2 years) of hospitalization for heart failure in patients assigned to saxagliptin. 20 This report explores further the observation surrounding hospitalizations for heart failure by examining baseline risk factors associated with an increased risk of hospitalizations for heart failure, the timing of hospitalizations and the risk of recurrent events, and the association between baseline levels of natriuretic peptides and future hospitalizations for heart failure events. Methods Study Design and Oversight As described previously, 19 the SAVOR-TIMI 53 trial was a multicenter, randomized, double-blind, placebo-controlled trial that randomized patients at 788 sites in 26 countries from May 2010 to December 2011 with type 2 DM, hemoglobin A1c between 6.5% and <12.0% within 6 months of randomization, and either a history of established cardiovascular disease or multiple risk factors for vascular disease to receive either 5 mg of saxagliptin daily (or 2.5 mg daily in patients with an estimated glomerular filtration rate [egfr] of 50 ml/min) or matching placebo. The full eligibility criteria and analysis plan have been reported previously. 19,20 Written informed consent was obtained from all patients. The relevant ethics committees at all participating centers approved the protocol. End Points A clinical events committee, unaware of the study-group assignments, adjudicated all components of the primary and secondary composite efficacy end points. The primary end point was a composite of cardiovascular death, myocardial infarction, or ischemic stroke. The secondary end point included the primary composite end point together with hospitalization for heart failure, coronary revascularization, or unstable angina. Heart failure requiring hospitalization was defined as an event that required hospitalization to an inpatient unit or a visit to an emergency department that resulted in at least a 12 hour stay with clinical manifestations of heart failure, including 1 of the following signs or symptoms: new or worsening dyspnea, orthopnea, paroxysmal nocturnal dyspnea, edema, pulmonary basilar crackles, jugular venous distension, new or worsening third heart sound or gallop rhythm, or radiological evidence of worsening heart failure, together with additional or increased therapy that included initiation of intravenous diuretic, inotropic, or vasodilator therapy, uptitration of intravenous therapy, if already on therapy, initiation of mechanical or surgical support/intervention, or the use of ultrafiltration, hemofiltration, or dialysis that is specifically directed at treatment of heart failure. 19,20 The end point definitions were developed to be consistent with the draft Standardized Definitions for End Point Events in Cardiovascular Trials created by an initiative from the Food and Drug Administration. 21 History of previous heart failure was a predefined subgroup based on medical history obtained at randomization. N-terminal pro-b type natriuretic peptide (NT-proBNP) was measured in patients (74.6% of the overall trial) at randomization and in a randomly selected subset of patients at 2 years or the end of treatment, whichever was earlier. Blood samples were collected in serum separator plastic tubes and then centrifuged and stored frozen in aliquots at 20 to 80 C at the enrolling site until shipped to the Biomarker Research/TIMI Clinical Trials Laboratory (Boston, MA), where they were maintained at 80 C. Serum NT-proBNP concentrations were measured at the first thaw using a sandwich immunoassay (probnp II; Roche Diagnostics, Indianapolis, IN). The analytic range extends from 5 to pg/ml. The reported within-run coefficient of variation was 4.2% at a level of 44 pg/ml and 2.7% at a level of pg/ml. Plasma high-sensitivity troponin T was measured with an electrochemiluminescent immunoassay assay (Roche Diagnostics). The lower limit of detection of the assay is μg/l. 22 Statistical Analysis All analyses were conducted on an intention-to-treat basis among patients who underwent randomization. Categorical variables were compared using χ 2 tests and continuous variable with either a t test or Wilcoxon rank-sum test, as appropriate. Events rates are presented as 2-year Kaplan-Meier estimates. The relative risks of hospitalization for heart failure between saxagliptin and placebo were examined using an unadjusted Cox proportional hazards model stratified by baseline renal impairment category and baseline cardiovascular risk group and with treatment as a model term. A post hoc Bonferroni correction was applied for the multiple comparisons of each composite of the secondary end point (n=6), providing a P value of for the comparison of saxagliptin versus placebo. Recurrent events analyses accounted for multiple hospitalizations for heart failure under a counting process assumption based on the method described by Anderson and Gill. 23 The risk of rehospitalization for heart failure after an initial hospitalization for heart failure was examined using the Prentice-Williams-Peterson Gap Time model. 24 Multivariable modeling for the risk of hospitalization for heart failure in the overall population was developed by first examining univariate associations and then with the backward elimination method to reduce the pool of covariates base on a P value of <0.05. It was refined further to include clinically meaningful covariates in the final model. The hazard ratio (HR) and 95% confidence interval (CI) were reported from landmark analyses based on subject-level censoring at 6 and 12 months that excluded subjects who did not have enough follow-up time or experienced the event before the landmarked time. A time-varying coefficient model was fitted as described by Gray 25 to evaluate any heterogeneity between saxagliptin and heart failure over time. Results There were a total of 741 hospitalizations for heart failure over a median follow-up of 2.1 years in 517 patients across both

3 Scirica et al Saxagliptin and Heart Failure 1581 treatment groups. A substantial proportion of patients hospitalized for heart failure, regardless of the treatment assignment, were subsequently readmitted for recurrent heart failure episodes (n=137, 26.5%) or died (n=135, 26.1%), mostly as a result of cardiovascular causes (n=121, 89.6% of deaths). A comparison of the baseline variables in the patients hospitalized for heart failure compared with those who were not are shown in Table 1. The former were older, more likely to be men, and have a history of heart failure, chronic kidney disease, established cardiovascular disease, including previous myocardial infarction, or coronary revascularization. In addition, they were more likely to be treated with aspirin, β-blockers, diuretics, statins, inhibitors of the renin-angiotensin system, and insulin and less likely to be treated with metformin, sulfonylureas, or thiazolidinediones. Risk Factors for Hospitalization for Heart Failure in the Overall Population Unadjusted HRs for the risk of hospitalization for heart failure by baseline characteristics are presented in Table 1. In multivariable analysis, the strongest association with hospitalization for heart failure (as assessed by χ 2 test) regardless of treatment assignment was previous heart failure and 2 markers of renal disease, egfr and the albumin/creatinine ratio (Table 2). There was a stepwise increase in the risk of hospitalization for heart failure in patients who had 0, 1, or 2 of the risk factors of history of heart failure or an egfr 60 m/min: 1.2% with 0 risk factors (n=10 418, 63.2%) (referent), 5.3% with 1 risk factor (n=5188, 31.5%; HR, 4.48; 95% CI, ; P<0.001), and 14.3% with 2 risk factors (n=886, 5.4%; HR, 13.51; 95% CI, ; P<0.001). When randomization is added to this model, the risk associated with saxagliptin was consistent with the overall trial (HR, 1.29; 95% CI, ). The c-statistic for the model with egfr and previous heart failure alone was 0.74, which increased to 0.81 with the addition of all clinical variables in Table 2. Saxagliptin and Hospitalization for Heart Failure Over 2 years of follow-up, more patients in the saxagliptin group (289 of 8280, 3.5%) were hospitalized for heart failure compared with the placebo group (228 of 8212, 2.8%; HR, 1.27; 95% CI, ; P=0.007). The corresponding rates at 6 months were 1.1% versus 0.6% (HR, 1.80; 95% CI, ; P=0.001) and 1.9% versus 1.3% (HR, 1.46; 95% CI, ; P=0.002) at 12 months (Figure 1). The rates for hospitalization for heart failure based on investigator reported events were consistent with the adjudicated results (see online-only Data Supplement Table I). Based on landmark analysis beginning at 6 and 12 months, the risk of hospitalization for heart failure after patients were assigned to saxagliptin was similar to placebo (2.4% versus 2.1%; HR, 1.11; 95% CI, ; P=0.31 beginning at 6 months; and 1.7% versus 1.5%; HR, 1.09; 95% CI, ; P=0.51 at 12 months). To examine the attenuating effects of saxagliptin on the risk of hospitalization as a result of heart failure over time, a time-varying coefficients model was developed (P value for a time time varying interaction term = 0.017). The lower CI of the log HR crosses 0 at?314 days, suggesting that the risk of hospitalization for heart failure with saxagliptin subsided at 10 to 11 months after randomization (see online-only Data Supplement Figure I). When the landmark analyses were restricted to patients taking the study drug, the corresponding risks were similar to the intent-to-treat analysis: from randomization to 12 months (HR, 1.52; 95% CI, ; P=0.0015) and from 12 months onward (HR, 1.05; 95% CI, ; P=0.73). When analyzing both first and recurrent events combined, 413 total events occurred in the saxagliptin group compared with 328 events in the placebo group (HR, 1.26; 95% CI, ; P=0.04). The risks for rehospitalizations for heart failure after an initial hospitalization for heart failure were similar in both groups (hospitalization for heart failure: 80, 27.7% saxagliptin versus 57, 25.0% for placebo; HR, 1.06; 95% CI, ; P=0.73; and mortality: 26.3% versus 25.9% placebo; HR, 1.01; 95% CI, ; P=0.95). The median length of stay for the initial adjudicated hospitalization for heart failure was 7.0 days in both treatment groups. The most common treatment was intravenous diuretics (88% in each group). Vasodilator therapy was used in 6.2% for saxagliptin subjects and in 7.9% for placebo subjects. Ultrafiltration/hemodialysis (2.8% and 2.2% for saxagliptin and placebo, respectively) and advanced hemodynamic support (1.7% and 0.9% for saxagliptin and placebo, respectively) were used infrequently. At 1 year, weights were similar in the overall saxagliptin and placebo groups (87.6±18.4 versus 87.9±19.4 kg; P=0.87) and in patients with a history of heart failure (91.5±19.5 versus 92.8±20.5 kg; P=0.27) or egfr 60 ml/min (87.1±18.8 versus 87.9±19.7 kg; P=0.47). Treatment with saxagliptin did not result in clinically detectable fluid overload as reflected by similar rates of adverse event reports of edema (45 for saxagliptin versus 46 for placebo) and peripheral edema (347 for saxagliptin versus 352 for placebo). Although the absolute rate of hospitalization for heart failure varied considerably among the predefined subgroups and in patients with and without previous heart failure, the relative risk in patients treated with saxagliptin was similar across subgroups (see online-only Data Supplement Figure II). Consequently, the absolute rates of heart failure difference between saxagliptin and placebo varied according to the overall risk within each subgroup. This pattern was reinforced when examining the relative and absolute risk differences in patients with the 2 clinical risk factors that most greatly increased the risk of hospitalization for heart failure, egfr 60 ml/min and a previous history of heart failure (Figure 2). For example, in patients with neither of these risk factors, the absolute risk difference over 2 years between saxagliptin and placebo was 0.3% compared with 1.7% in patients with both risk factors. The number of excess hospitalizations for heart failure per year if 1000 patients were treated with saxagliptin ranged from 0 in patients with no risk factors to 8 in patients with both risk factors. Baseline NT-proBNP and Heart Failure The median NT-proBNP in the patients in whom it was measured at baseline was 141 pg/ml (interquartile ratio, ), and it was similar in patients assigned to saxagliptin (143 pg/ml; interquartile ratio, ) and placebo (139 pg/ml; interquartile ratio, ). There was a stepwise increased risk of hospitalization for heart failure with higher quartiles of baseline NT-proBNP (Figure 3). When added to

4 1582 Circulation October 28, 2014 Table 1. Baseline Characteristics Characteristic Hospitalization for Heart Failure (n = 517) No Hospitalization for Heart Failure (n = ) P Unadjusted HR (95% CI) Demographic characteristics Age (y), median (IQR) 68.0 ( ) 65.0 ( ) < ( )* Age 75 y, n(%) 137 (26.5%) 2193 (13.7%) < ( ) Male, n(%) 377 (72.9%) (66.7%) < ( ) Race <0.01 White, n(%) 424 (82.0%) (75.0%) 1.45 ( )* Black, n(%) 32 (6.2%) 536 (3.4%) Asian, n(%) 35(6.8%) 1745(10.9%) Multiracial, n(%) 21(4.1%) 1505(9.4%) Other, n(%) 5 (1.0%) 206 (1.3) Hispanic, n(%) 68 (13.2%) 3473 (21.7%) < ( ) Weight (kg), median (IQR) 89.9 ( ) 86.0 ( ) < ( )* Weight >80 kg, n(%) 360 (69.8%) (63.9%) ( ) Body mass index (kg/m 2 ), median (IQR) 31.5 ( ) 30.4 ( ) < ( )* Body mass index 30, n(%) 299 (57.9%) 8517 (53.4%) ( ) Duration of diabetes, median (IQR) 12.7 ( ) 10.3 ( ) < ( )* Established atherosclerotic disease, n(%) 470 (90.9%) (78.2%) < ( ) Hypertension, n(%) 421 (81.4%) (81.8%) ( ) Dyslipidemia, n(%) 411 (79.5%) (70.9%) < ( ) Previous myocardial infarction, n(%) 282 (54.5%) 5955 (37.3%) < ( ) Previous heart failure, n(%) 226 (43.7%) 1879 (11.8%) < ( ) Previous coronary revascularization, n(%) 296 (57.3%) 6827 (42.7%) < ( ) Glycohemoglobin, median (IQR) 7.7 ( ) 7.6 ( ) ( )* Glycohemoglobin, n(%) 0.44 <6.5% 33 (6.5%) 1230 (7.8%) Ref. 6.5% to <7.0% 81 (15.9%) 2775 (17.7%) 1.09 ( ) 7.0% to <8.0% 187 (36.7%) 5229 (33.3%) 1.35 ( ) 8.0% to <9.0% 98 (19.2%) 3041 (19.4%) 1.24 ( ) 9.0% 111 (21.8%) 3414 (21.8%) 1.26 ( ) Fasting serum glucose (mg/dl), median (IQR) ( ) ( ) ( ) Estimated glomerular filtration rate (ml/min), median (IQR) 55.7 ( ) 72.1 ( ) < ( ) Estimated glomerular filtration rate, n(%) <0.01 <30 ml/min 44 (8.5%) 288 (1.8%) 7.26 ( ) 30 to 60 ml/min 248 (48.0%) 4275 (26.8%) 2.98 ( ) >60 ml/min 225 (43.5%) (71.4%) Ref. Albumin/creatinine ratio (mg/mmol), median (IQR) 7.1 ( ) 1.8 ( ) < ( ) Albumin/creatinine ratio, n(%) <0.01 <3.4 mg/mmol 182 (36.9%) 9514 (62.3%) Ref. 3.4 to 33.9 mg/mmol 177 (35.9%) 4249 (27.8%) 2.20 ( ) >33.9 mg/mmol 134 (27.2%) 1504 (9.9%) 4.70 ( ) Baseline cardiovascular medications, n(%) Aspirin 416 (80.5%) (75.0%) ( ) β-blockers 417 (80.7%) 9745 (61.0%) < ( ) ACE inhibitors 303 (58.6%) 8637 (54.1%) ( ) Angiotensin receptor blockers 142 (27.5%) 4453 (27.9%) ( ) Diuretics 386 (74.7%) 6812 (42.6%) < ( ) Statins 429 (83.0%) (78.2%) ( ) Calcium antagonists 174 (33.7%) 5204 (32.6%) ( ) (Continued )

5 Scirica et al Saxagliptin and Heart Failure 1583 Table 1. Continued Characteristic Hospitalization for Heart Failure (n = 517) No Hospitalization for Heart Failure (n = ) P Unadjusted HR (95% CI) Baseline anti-hyperglycemic medications, n(%) Metformin 270 (52.2%) (69.8%) < ( ) Sulfonylurea 156 (30.2%) 6441 (40.3%) < ( ) Thiazolidinediones 18 (3.5%) 954 (6.0%) ( ) Insulin 302 (58.4%) 6484 (40.6%) < ( ) Hazard ratio, CI and P value for risk of event for subjects with characteristic relative to base category. Data on race/ethnicity were entered by study coordinators based on patient self-identification. ACE indicates angiotensin converting enzyme; CI, confidence interval; HR, hazard ratio; and IQR, interquartile range. *Per unit of measurement defined in the 1st column. HR per 10 U increase for fasting glucose and egfr. HR per 1 U increase for log albumin:creatinine ration. the clinical multivariable model, quartile 4 of NT-proBNP level was the factor most strongly associated with the risk of hospitalization for heart failure (HR, 5.47; 95% CI, ; P<0.001; see online-only Data Supplement Table II). The c-statistic increased from 0.81 to 0.85 (P<0.001) when NT-proBNP was added to the clinical variables in Table 2. In the cohort with the highest quartile of NT-proBNP, the risk of hospitalization for heart failure in patients treated with saxagliptin was similar to that observed in the overall trial (HR, 1.31; 95% CI, ; P=0.02). Although there was no evidence of heterogeneity between levels of NT-proBNP, treatment with saxagliptin, and hospitalization for heart failure (P for interaction=0.46), the absolute risk excess for heart failure with saxagliptin was greatest in the highest NT-proBNP quartile (2.1%) compared with quartiles 1 (0.0%), 2 (0.7%), and 3 (0.2%) (Figure 3). Similar results were seen when evaluating NT-proBNP according to deciles or an established dichotomous cut point (see online-only Data Supplement Figure III). The addition of the highest quartile of baseline NT-proBNP to the risk factors of history of heart failure and egfr further stratified patients according to baseline risk of hospitalization for heart failure regardless of treatment assignment (see online-only Data Supplement Figure IV). The addition of the NT-proBNP to the 2 clinical variables increased the c-statistic from 0.74 to 0.82 (P<0.001). Among the lowest quartiles of NT-proBNP, most patients had none of the aforementioned risk factors for heart failure (n=6599, 53.6%), a correspondingly low rate of heart failure, and no treatment difference between saxagliptin or placebo (0.6% versus 0.6%; HR, 0.84; 95% CI, ; P=0.56), whereas in patients with NT-proBNP in the top quartile and 2 risk factors, the absolute risk difference was 1.6% (see online-only Data Supplement Figure V). The baseline concentration of NT-proBNP was a particularly strong risk factor for hospitalization for heart failure. Even among patients with normal renal function and no reported history of heart failure, a level of NT-proBNP in quartile 4 (>332 pg/ml) was associated with a higher risk of hospitalization for heart failure (4.3% versus 0.6% with lower levels of NT-proBNP; see online-only Data Supplement Figure IV) and a correspondingly higher absolute risk with saxagliptin compared with placebo (5.35% versus 3.33%; P=0.083; see online-only Data Supplement Figure V). Change in NT-proBNP, High-Sensitivity Troponin T, and C-Reactive Protein In patients with both baseline and 2 years or end-oftreatment assessments, median NT-proBNP (n=2026) Table 2. Multivariable Analysis Evaluating the Relationship between Baseline Clinical Characteristics and Risk for Hospitalization for Heart Failure in the Overall SAVOR-TIMI 53 Population n χ 2 Hazard Ratio Adjusted 95% Confidence Intervals Previous heart failure <0.01 Albumin/creatinine ratio >33.9 mg/mmol <0.01 Albumin/creatinine ratio 3.4 to 33.9 mg/mmol <0.01 Estimated glomerular filtration rate 60 ml/min <0.01 Age 75 y <0.01 Previous myocardial infarction <0.01 Non-Hispanic <0.01 Established cardiovascualr disease <0.01 Saxagliptin Female Dyslipidemia SAVOR-TIMI 53 indicates Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus Thrombolysis in Myocardial Infarction 53. P

6 1584 Circulation October 28, 2014 Hospitalization for Heart Failure (%) Placebo Saxagliptin increased in patients treated with placebo ( pg/ml; P<0.001) and saxagliptin ( pg/ml; P=0.001), with a slightly greater increase in placebo (median change, 10 versus 4 pg/ml; P=0.001). The pattern of change was similar in patients with previous heart failure (n=217) in the placebo ( pg/ml; P=0.149; n=100) and saxagliptin ( pg/ml; P=0.238; n=117; median change, 15 versus 4 pg/ml; P=0.70) groups, and patients without previous heart failure (n=1809) in the placebo ( pg/ml; P<0.001) and saxagliptin ( pg/ml; P=0.003; median change, 10 versus 4 pg/ml; P<0.001) groups. Similar results were observed after excluding patients who experienced any of the primary or secondary end points (data not shown). In the subset of patients with baseline and follow-up biomarkers, there were no differences between placebo and saxagliptin in the median change in concentrations from baseline to 2 years or end of treatment of high-sensitivity troponin T (n=1355; 0.71 versus 0.52 ng/l; P=0.067) or C-reactive protein (n=513; 0.08 versus 0.15 μg/ml; P=0.75). Patients with Previous Heart Failure A total of 2105 patients (12.8%) reported previous heart failure at baseline. They were older and more likely to have established cardiovascular disease and renal impairment, although Hospitalization for HF (2yr KM %) 4% 3% 2% 1% 25% 20% 15% 10% 5% 0% # of excess HHF events/ 1000 pt-year 0.6% 0% HR 1.80 ( ) P=0.001 ARD 0.3% HR 1.16 ( ) P= % 1.8% egfr > 60ml/min ARD 1.5% HR 1.36 ( ) P= % 5.3% egfr =<60 ml/min Saxagliptin ARD 0.6% HR 1.30 ( ) P= % 1.7% No Prior Heart Failure ARD 1.5% HR 1.23 ( ) P= % 10.2% Prior Heart Failure Placebo ARD 0.3% HR 1.15 ( ) P= % 1.0% 0 Risk Factors ARD 1.4% HR 1.35 ( ) P= % 4.6% there was no difference in the duration of DM and only a small difference in baseline hemoglobin A1c (7.7% in previous heart failure versus 7.6% in no previous heart failure; P=0.08). More patients with previous heart failure were on aspirin, statins, β-blockers, and insulin, but fewer were on metformin or sulfonylureas (see online-only Data Supplement Table III). Patients with previous heart failure were at increased risk of all cardiovascular events compared with those without previous heart failure. The relative effect of saxagliptin versus placebo was similar in patients with and without previous heart failure for the primary and secondary composite end points, as well as all-cause mortality. However, the absolute increase in the rate of hospitalization for heart failure with saxagliptin was 1.5% in patients with previous heart failure compared with 0.6% in patients without previous heart failure (P for interaction=0.67). The rates of cardiovascular events, as well as hypoglycemia, and adverse events are presented in the online-only Data Supplement Table IV. We further evaluated the risk of hospitalization for heart failure in patients with previous heart failure by New York Heart Association Functional Classification (see online-only Data Supplement Table V). Similar to the overall trial, the primary and secondary endpoints were balanced between treatment arms in patients with higher NHYA classification (Data not shown). Discussion In patients with established cardiovascular disease or multiple cardiovascular risk factors, the DPP-4 inhibitor saxagliptin, when compared with placebo, increased the risk of hospitalization for heart failure, in particular during the first 12 months of therapy. Several clinical features easily identified patients at highest absolute risk of heart failure, regardless of treatment assignment. Although there were no individual predefined subgroups in which the relative risk associated with saxagliptin treatment was particularly high or low, the incremental risk with saxagliptin was greatest in patients at a high overall risk of heart failure (ie, history of heart failure, impaired renal function, or elevated baseline levels of NT-proBNP) and correspondingly small in patients at lower risk. Therefore, a combination of clinical factors and biomarkers identify a population of patients with DM in whom the excess risk of hospitalization for heart failure with saxagliptin treatment is greatest, while ARD 1.7% HR 1.22 ( ) P= % 13.4% 1 Risk Factor 2 Risk Factors N= ,418 5, % Saxagliptin HR 1.46 ( ) P= % 1.3% Placebo Days from Randomization HR 1.27 ( ) P= % 2.8% Figure 1. Kaplan-Meier failure estimates of hospitalization for heart failure according to treatment with saxagliptin versus placebo. Kaplan-Meier estimates and corresponding hazard ratios (HR) are presented at 6 months, 12 months, and 2 years Figure 2. Risk of hospitalization for heart failure with saxagliptin or placebo in patients with and without baseline risk factors (egfr <60 ml/ min or history of previous heart failure). ARD indicates absolute risk difference; egfr, estimated glomerular filtration rate; HF, heart failure; HR, hazard ratio; KM, Kaplan-Meier; and # excess HHR events/1000 pt-years, the number of excess hospitalizations for heart failure in patients treated with saxagliptin versus placebo per 1000 patient-years.

7 Scirica et al Saxagliptin and Heart Failure 1585 Hosp. for Heart Failure (%) 10% 8% 6% 4% 2% 0% # of excess HHF events/ 1000 pt-year Saxagliptin p for interaction=0.46 ARD 0.7% ARD 0% HR 1.82 HR 1.04 ( ) ( ) P=0.12 P= % 0.1% 0.1% 0.4% Placebo ARD 0.2% HR 0.94 ( ) P= % 2.0% ARD 2.1% HR 1.31 ( ) P= % 8.9% Q1 Q2 Q3 Q4 (5-64) (65-141) ( ) (333-46,627) N= Quartiles of NT-proBNP (pg/ml) Figure 3. Risk of hospitalization for heart failure with saxagliptin or placebo according to baseline quartile of NT-proBNP (picograms per milliliter). ARD indicates absolute risk difference; HR, hazard ratio; NT-proBNP, N-terminal pro B-type natriuretic peptide; Q, quartile; and # excess HHR events/1000 pt-years, the number of excess hospitalizations for heart failure in patients treated with saxagliptin versus placebo per 1000 patient-years. conversely identifying a larger population without those features in whom the absolute risk is small. Based on the mechanism of action of saxagliptin and the accumulated preclinical and clinical data of DPP-4 antagonists, the observation of a higher incidence of hospitalization for heart failure in patients treated with saxagliptin in the SAVOR-TIMI 53 trial was unexpected and requires confirmation with several ongoing cardiovascular outcomes trials of DPP-4 inhibitors and glucagon-like peptide-1 agonists. 26,27 There was no signal of fluid retention, weight gain, or heart failure with saxagliptin in the Phase 2 and Phase 3 development program. 28 Moreover, previous preclinical data and small studies with other incretin-based agents suggested potential improvement of left ventricular function, 26 but these data are inconsistent. Two preliminary reports of other studies in patients with DDP-4 inhibitors produced unanticipated findings that highlight potential mechanisms by which treatment with DPP-4 inhibitors could exacerbate heart failure. In 1 study in patients with reduced left ventricular function, after 12 months of therapy, the DPP-4 inhibitor vildagliptin increased left ventricular end diastolic volumes. 27 In another 6-week trial, several DPP-4 inhibitors worsened endothelial function as assessed by flow-mediated dilation, independent of glucagon-like peptide-1 levels or DPP-4 activity. 29 Regardless, no clear mechanism of action to explain the increased risk of hospitalization for heart failure in the SAVOR-TIMI 53 trial could be identified. Although unexpected, the incremental risk of heart failure hospitalization observed with saxagliptin is likely valid, given the large number of events and the prespecification of heart failure hospitalizations as a component of the secondary end point, together with central blinded adjudication. However, the observation of an increased risk of hospitalization for heart failure with saxagliptin must be taken in the context of multiple testing and the risk of a false-positive result, although there was a statistically significant difference between the 2 groups after post hoc adjustment for multiple comparisons. Left ventricular function was not captured at baseline, and therefore the temporal changes, even in patients with previous heart failure, cannot be evaluated. The Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care in Patients with Type 2 Diabetes Mellitus and Acute Coronary Syndrome (EXAMINE) trial that evaluated the DPP-4 inhibitor alogliptin versus placebo also identified a numerically similar trend for increased risk of hospitalization for heart failure, but the overall number of events was smaller (106 [3.9%] versus 89 [3.3%]; HR, 1.19; 95% CI, ; P=0.22). 30 When the 712 first hospitalization for heart failure events from both studies are combined, the overall odds of hospitalization for heart failure with DPP-4 inhibition is 1.24 (95% CI, ; P=0.004; see online-only Data Supplement Figure VI). An increased risk of heart failure has been observed in some studies of other antihyperglycemic agents, including sulfonylureas, 13 thiazolidinediones, 7 9,11 and dual peroxisome proliferator-activated receptor α/γ agonists, 10,31 with divergent results in trials of intensive glucose management. 32,33 The lack of adequate safety data on antihyperglycemic agents in heart failure is highlighted by the example of metformin, which for many years was contraindicated in patients with heart failure, until observational data demonstrated an acceptable safety profile. 15,16 Patients with both DM and heart failure are relatively under-represented in clinical trials that either excluded them or enrolled <5% of the population. 37,38 Therefore, the 2105 patients with previous heart failure in the SAVOR-TIMI 53 trial represent 1 of the largest cohorts of such patients with DM studied. 1 There are presently no known mechanisms by which DPP-4 inhibition could precipitate heart failure. The hemodynamic effects of glycemic modulation in myocardium accustomed to years of hyperglycemia are unknown and could potentially exacerbate cardiac dysfunction because glycemic changes may unfavorably alter the balance of free fatty acid oxidation and glycolysis. 39 Intensive glycemic control increased the risk of heart failure in 1 of the large glucose-lowering trials 40 but not in 2 others. 36,38 In contrast to the thiazolidinediones, there is no signal of volume overload observed in the SAVOR- TIMI 53 trial with saxagliptin, nor did saxagliptin raise levels of NT-proBNP. However, interestingly, treatment with the glitazones increases levels of natriuretic peptides in some studies 41 but not all studies. 7 The cardiovascular consequences of DPP-4 inhibition on other peptide substrates, such as natriuretic peptides or bradykinins, are also unknown, nor was there evidence of direct myocardial toxicity with saxagliptin as reflected by the similar change in concentrations of high-sensitivity troponin T and high-sensitivity C-reactive protein between treatment groups. With the possible exception of metformin 15,16 and insulin, 42 most reported studies to date evaluating effects on heart failure of specific glucose-lowering medications either increased the risk of heart failure or were insufficiently powered and therefore often discordant. 33 Moreover, the trials of metformin and insulin by design enrolled patients

8 1586 Circulation October 28, 2014 with recently diagnosed DM in whom glycemic modulation may be better tolerated by the myocardium. In this wellpowered study of patients at high cardiovascular risk, we found that saxagliptin did not increase the overall risk of the primary or secondary cardiovascular composite end points, even in those with previous heart failure. However, patients treated with saxagliptin had an increased risk of hospitalization for heart failure an event that itself was associated with a high subsequent mortality and in particular in patients at the highest absolute risk of heart failure. The decision to choose 1 antihyperglycemic agent versus another must balance the benefit in reducing microvascular complications via improved glycemic control together with potential adverse events, such as hypoglycemia and heart failure. The results from contemporary large cardiovascular outcome studies in patients with DM across the spectrum of classes of antihyperglycemic agents will greatly improve our evidence-based approach to treatment of these 2 commonly coexistent conditions. Acknowledgments The authors would like to recognize the Executive Committee: Eugene Braunwald (Study Chair), Deepak L. Bhatt (Co-Principal Investigator), Itamar Raz (Co-Principal Investigator), Jaime A. Davidson, Robert Frederich (non-voting), Boaz Hirshberg (non-voting), and Ph. Gabriel Steg. Sources of Funding SAVOR-TIMI 53 was funded by AstraZeneca and Bristol-Myers Squibb. Biomarker reagents supported by Roche Diagnostics. Disclosures Dr Scirica reports research grants via the TIMI Study and Brigham and Women s Hospital from AstraZeneca and Bristol- Myers Squibb, Daichi-Sankyo, GlaxoSmithKline, Johnson & Johnson, Bayer Healthcare, Gilead, Eisai, and Merck; and consulting fees from AstraZeneca, Gilead, Lexicon, Arena, Eisai, St. Jude s Medical, Bristol-Myers Squibb, Forest Pharmaceuticals, Boston Clinical Research Institute, Decision Resources, University of Calgary, Elsevier Practice Update Cardiology, and Forest Pharmaceuticals. Dr Braunwald reports grants from Astra Zeneca, Bristol-Myers Squibb while this study was conducted, Johnson & Johnson, Merck, Sanofi Aventis, Daiichi Sankyo, GlaxoSmithKline, Beckman Coulter, Roche Diagnostics, Pfizer, Eli Lilly, and Duke University; and, outside the submitted work, personal fees from Eli Lilly, Merck, CVRx, CV Therapeutics (now Gilead), Daiichi Sankyo, Menarini International, Medscape, Bayer, Genzyme, Medicines Company, and Sanofi Aventis. Dr Raz reports grants from AstraZeneca and Bristol-Myers Squibb while this study was conducted; scientific board membership from Novo Nordisk, MSD, Eli Lilly, Sanofi, Medscape, Andromeda, and Insuline; payment for lectures including service on speakers bureaus for lectures from Eli Lilly, Novo Nordisk, Johnson & Johnson, Sanofi, MSD, and Novartis; and stock options in Insuline. Dr Morrow reports grants from AstraZeneca while this study was conducted; consultancy fees from Abbott Laboratories, BG Medicine, Critical Diagnostics, Daiichi Sankyo, Genetech, Gilead, Instrumentation Laboratories, Johnson & Johnson, Konica Minolta, Merck, Novartis, Provenchio, Roche Diagnostics, and Servier; and grants from Abbott Laboratories, Beckman Coulter, BG Medicine, Critical Diagnostics, Bristol-Myers Squibb, Daiichi Sankyo, GlaxoSmithKline, Merck, Novartis, Roche Diagnostics, Sanofi Aventis, and Gilead. Dr Jarolim reports grants from Abbott, AstraZeneca, Daiichi Sankyo, Merck, Roche Diagnostics, and Waters Technologies; and consultancy fees from Biosystems and Quanterix. Dr Mosenzon reports grants from AstraZeneca and Bristol-Myers Squibb while this study was conducted; consulting fees from AstraZeneca and Bristol- Myers Squibb; support for travel to meetings for the study from AstraZeneca and Bristol-Myers Squibb; scientific advisory board membership from Novo Nordisk, Eli Lilly, Sanofi, Norvartis; and speakers bureaus for Novo Nordisk, Eli Lilly, Sanofi, Norvartis, Merck Sharpe & Dohme. Drs Pollack and Hirshberg report employment by AstraZeneca and having stock/stock options in AstraZeneca. Dr Frederich reports employment by Bristol-Myers Squibb and having stock/stock options in Bristol-Myers Squibb. Dr Lewis reports grants from AstraZeneca while this study was conducted; scientific advisory board membership from MSD and AstraZeneca; and grants from Bayer Healthcare, Amylin, Amgen, GlaxoSmithKline, MSD, Eli Lilly, and Sanofi. Dr McGuire reports grants from Brigham and Women s Hospital; personal fees from Brigham and Women s Hospital while this study was conducted; personal fees from Boehringer Ingelheim, Janssen Research and Development, Sanofi Aventis Groupe, Genentech, Merck Sharp & Dohme, Medscape Cardiology, Pri-Med Institute, Brigham and Women s Hospital, Duke Clinical Research Institute, Cleveland Clinic Coordinating Center for Clinical Research, University of Oxford, Daiichi Sankyo, Eli Lilly, Novo Nordisk, F. Hoffmann La Roche, Axio Research, Premier Research, INC Research, GlaxoSmithKline, Takeda Pharmaceuticals North America, Bristol-Myers Squibb, Eisai, Omthera, and Regeneron outside the submitted work; and nonfinancial support from Gilead Sciences. Dr Davidson reports personal fees from the TIMI Group while this study was conducted. Dr Steg reports personal fees from AstraZeneca while this study was conducted; personal fees from Amarin, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Daiichi Sankyo, GlaxoSmithKline, Eli Lilly, Merck Sharpe &Dohme, Novartis, Otsuka, Pfizer, Roche, Medicines Company, and Vivus outside the submitted work; and grants and personal fees from Sanofi and Servier outside the submitted work. Dr Bhatt discloses the following relationships: Advisory Boards for Elsevier Practice Update Cardiology, Medscape Cardiology, and Regado Biosciences; Boards of Directors for Boston Veterans Affairs Research Institute, and Society of Cardiovascular Patient Care; Chair for American Heart Association Get With The Guidelines Steering Committee; Data Monitoring Committees for Duke Clinical Research Institute, Harvard Clinical Research Institute, Mayo Clinic, and Population Health Research Institute; Honoraria for American College of Cardiology (Editor, Clinical Trials, Cardiosource), Belvoir Publications (Editor in Chief, Harvard Heart Letter), Duke Clinical Research Institute (clinical trial steering committees), Harvard Clinical Research Institute (clinical trial steering committee), HMP Communications (Editor in Chief, Journal of Invasive Cardiology), Population Health Research Institute (clinical trial steering committee), Slack Publications (Chief Medical Editor, Cardiology Today s Intervention), WebMD (CME steering committees); Other: Clinical Cardiology (Deputy Editor), Journal of the American College of Cardiology (Section Editor, Pharmacology); Research Grants from Amarin, AstraZeneca, Bristol-Myers Squibb, Eisai, Ethicon, Medtronic, Roche, Sanofi Aventis, Medicines Company; and unfunded research with FlowCo, PLx Pharma, and Takeda. Drs Cavender and Udell and Ms. Umez-Eronini report no conflicts of interest. References 1. McMurray JJV, Gerstein HC, Holman RR, Pfeffer MA. Heart failure: a cardiovascular outcome in diabetes that can no longer be ignored. Lancet Diabetes Endocrinol. March 13, doi: /s (14) Accessed August 27, Rodrigues B, Cam MC, McNeill JH. Myocardial substrate metabolism: implications for diabetic cardiomyopathy. J Mol Cell Cardiol. 1995;27: Opie LH, Yellon DM, Gersh BJ. Controversies in the cardiovascular management of type 2 diabetes. Heart. 2011;97:6 14.

9 Scirica et al Saxagliptin and Heart Failure Boudina S, Abel ED. Diabetic cardiomyopathy revisited. Circulation. 2007;115: Schilling JD, Mann DL. Diabetic cardiomyopathy: bench to bedside. Heart Fail Clin. 2012;8: Nichols GA, Hillier TA, Erbey JR, Brown JB. Congestive heart failure in type 2 diabetes: prevalence, incidence, and risk factors. Diabetes Care. 2001;24: Bhatt DL, Chew DP, Grines C, Mukherjee D, Leesar M, Gilchrist IC, Corbelli JC, Blankenship JC, Eres A, Steinhubl S, Tan WA, Resar JR, AlMahameed A, Abdel-Latif A, Wilson Tang WH, Tang HW, Brennan D, McErlean E, Hazen SL, Topol EJ. Peroxisome proliferator-activated receptor gamma agonists for the prevention of adverse events following percutaneous coronary revascularization results of the PPAR study. Am Heart J. 2007;154: Dormandy JA, Charbonnel B, Eckland DJ, Erdmann E, Massi-Benedetti M, Moules IK, Skene AM, Tan MH, Lefèbvre PJ, Murray GD, Standl E, Wilcox RG, Wilhelmsen L, Betteridge J, Birkeland K, Golay A, Heine RJ, Korányi L, Laakso M, Mokán M, Norkus A, Pirags V, Podar T, Scheen A, Scherbaum W, Schernthaner G, Schmitz O, Skrha J, Smith U, Taton J; PROactive investigators. Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive Study (PROspective pioglitazone Clinical Trial In macrovascular Events): a randomised controlled trial. Lancet. 2005;366: 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: Nissen SE, Wolski K, Topol EJ. Effect of muraglitazar on death and major adverse cardiovascular events in patients with type 2 diabetes mellitus. JAMA. 2005;294: Home PD, Pocock SJ, Beck-Nielsen H, Curtis PS, Gomis R, Hanefeld M, Jones NP, Komajda M, McMurray JJ; RECORD Study Team. Rosiglitazone evaluated for cardiovascular outcomes in oral agent combination therapy for type 2 diabetes (RECORD): a multicentre, randomised, open-label trial. Lancet. 2009;373: Eurich DT, Majumdar SR, McAlister FA, Tsuyuki RT, Johnson JA. Improved clinical outcomes associated with metformin in patients with diabetes and heart failure. Diabetes Care. 2005;28: Nichols GA, Koro CE, Gullion CM, Ephross SA, Brown JB. The incidence of congestive heart failure associated with antidiabetic therapies. Diabetes Metab Res Rev. 2005;21: Eurich DT, McAlister FA, Blackburn DF, Majumdar SR, Tsuyuki RT, Varney J, Johnson JA. Benefits and harms of antidiabetic agents in patients with diabetes and heart failure: systematic review. BMJ. 2007;335: Eurich DT, Weir DL, Majumdar SR, Tsuyuki RT, Johnson JA, Tjosvold L, Vanderloo SE, McAlister FA. Comparative safety and effectiveness of metformin in patients with diabetes mellitus and heart failure: systematic review of observational studies involving 34,000 patients. Circ Heart Fail. 2013;6: Roussel R, Travert F, Pasquet B, Wilson PW, Smith SC Jr, Goto S, Ravaud P, Marre M, Porath A, Bhatt DL, Steg PG; Reduction of Atherothrombosis for Continued Health (REACH) Registry Investigators. Metformin use and mortality among patients with diabetes and atherothrombosis. Arch Intern Med. 2010;170: Iribarren C, Karter AJ, Go AS, Ferrara A, Liu JY, Sidney S, Selby JV. Glycemic control and heart failure among adult patients with diabetes. Circulation. 2001;103: Inzucchi SE, McGuire DK. New drugs for the treatment of diabetes: part II: Incretin-based therapy and beyond. Circulation. 2008;117: Scirica BM, Bhatt DL, Braunwald E, Steg PG, Davidson J, Hirshberg B, Ohman P, Price DL, Chen R, Udell J, Raz I. The design and rationale of the saxagliptin assessment of vascular outcomes recorded in patients with diabetes mellitus-thrombolysis in myocardial infarction (SAVOR-TIMI) 53 study. Am Heart J. 2011;162: e Scirica BM, Bhatt DL, Braunwald E, Steg PG, Davidson J, Hirshberg B, Ohman P, Frederich R, Wiviott SD, Hoffman EB, Cavender MA, Udell JA, Desai NR, Mosenzon O, McGuire DK, Ray KK, Leiter LA, Raz I; SAVOR-TIMI 53 Steering Committee and Investigators. Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus. N Engl J Med. 2013;369: Hicks KA, Hung HMJ, Mahaffey KW, Mehran R, Nissen SE, Stockbridge NL, Targum SL, Temple R. Standardized definitions for end point events in cardiovascular trials meetings/meeting-brochures/parc-meeting-materials/february-3rd-2012/ articles/standardized%20definitions.pdf/view. 22. Giannitsis E, Becker M, Kurz K, Hess G, Zdunek D, Katus HA. Highsensitivity cardiac troponin T for early prediction of evolving non-stsegment elevation myocardial infarction in patients with suspected acute coronary syndrome and negative troponin results on admission. Clin Chem. 2010;56: Andersen PK, Gill RD. Cox s regression model for counting processes: a large sample study. Ann Statistics. 1982;10: Prentice RL, Williams BJ, Petersen V. On the regression analysis of multivariate failure time data. Biometrika 1981;68: Gray RJ. Flexible methods for analyzing survival data using splines, with applications to breast cancer prognosis. J Am Stat Assoc. 1992;87: Khan MA, Deaton C, Rutter MK, Neyses L, Mamas MA. Incretins as a novel therapeutic strategy in patients with diabetes and heart failure. Heart Fail Rev. 2013;18: McMurray JJ, Ponikowksi P, Bolli GB, Kozlovski P, Jhund P, Lewsey JD, Moeckel V, Lukashevich V, Kothny W, Krum H. Vildagliptin in Ventricular Dysfunction Diabetes Trial (VIVIDD). Eur J Heart Fail. 2013; Iqbal N, Parker A, Frederich R, Donovan M, Hirshberg B. Assessment of the cardiovascular safety of saxagliptin in patients with type 2 diabetes mellitus: pooled analysis of 20 clinical trials. Cardiovasc Diabetol. 2014;13: Ayaori M, Iwakami N, Uto-Kondo H, Sato H, Sasaki M, Komatsu T, Iizuka M, Takiguchi S, Yakushiji E, Nakaya K, Yogo M, Ogura M, Takase B, Murakami T, Ikewaki K. Dipeptidyl peptidase-4 inhibitors attenuate endothelial function as evaluated by flow-mediated vasodilatation in type 2 diabetic patients. J Am Heart Assoc. 2013;2:e White WB. Cardiovascular outcomes with alogliptin in patients with type 2 diabetes mellitus and recent acute coronary syndromes. European Association for the Study of Diabetes. Barcelona2013. April 3, Accessed September 16, Lincoff AM, Tardif JC, Schwartz GG, Nicholls SJ, Rydén L, Neal B, Malmberg K, Wedel H, Buse JB, Henry RR, Weichert A, Cannata R, Svensson A, Volz D, Grobbee DE; AleCardio Investigators. Effect of aleglitazar on cardiovascular outcomes after acute coronary syndrome in patients with type 2 diabetes mellitus: the AleCardio randomized clinical trial. JAMA. 2014;311: Boussageon R, Bejan-Angoulvant T, Saadatian-Elahi M, Lafont S, Bergeonneau C, Kassaï B, Erpeldinger S, Wright JM, Gueyffier F, Cornu C. Effect of intensive glucose lowering treatment on all cause mortality, cardiovascular death, and microvascular events in type 2 diabetes: metaanalysis of randomised controlled trials. BMJ. 2011;343:d Ray KK, Seshasai SR, Wijesuriya S, Sivakumaran R, Nethercott S, Preiss D, Erqou S, Sattar N. Effect of intensive control of glucose on cardiovascular outcomes and death in patients with diabetes mellitus: a meta-analysis of randomised controlled trials. Lancet. 2009;373: Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). UK Prospective Diabetes Study (UKPDS) Group. Lancet. 1998;352: Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet. 1998;352: Patel A, MacMahon S, Chalmers J, Neal B, Billot L, Woodward M, Marre M, Cooper M, Glasziou P, Grobbee D, Hamet P, Harrap S, Heller S, Liu L, Mancia G, Mogensen CE, Pan C, Poulter N, Rodgers A, Williams B, Bompoint S, de Galan BE, Joshi R, Travert F. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med. 2008;358: Gerstein HC, Miller ME, Byington RP, Goff DC Jr, Bigger JT, Buse JB, Cushman WC, Genuth S, Ismail-Beigi F, Grimm RH Jr, Probstfield JL, Simons-Morton DG, Friedewald WT. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med. 2008;358: Duckworth W, Abraira C, Moritz T, Reda D, Emanuele N, Reaven PD, Zieve FJ, Marks J, Davis SN, Hayward R, Warren SR, Goldman S, McCarren M, Vitek ME, Henderson WG, Huang GD; VADT Investigators. Glucose control and vascular complications in veterans with type 2 diabetes. N Engl J Med. 2009;360: Dutka DP, Pitt M, Pagano D, Mongillo M, Gathercole D, Bonser RS, Camici PG. Myocardial glucose transport and utilization in patients with type 2 diabetes mellitus, left ventricular dysfunction, and coronary artery disease. J Am Coll Cardiol. 2006;48: