14 Diabetes Care Volume 41, January 2018 EXPERT FORUM

Cardiovascular Outcomes Trials in William T. Cefalu,1 Sanjay Kaul,2 Hertzel C. Gerstein,3 Rury R. Holman,4 5 6 DIABETES CARE Type2Diabetes:WhereDoWeGo Bernard Zinman, Jay S. Skyler, Jennifer B. Green,7 John B. Buse,8 From Here? Reflections From a Silvio E. Inzucchi,9 Lawrence A. Leiter,10 Itamar Raz,11 Julio Rosenstock,12 and Diabetes Care Editors’ Expert Matthew C. Riddle13 Forum

Diabetes Care 2018;41:14–31 | https://doi.org/10.2337/dci17-0057

1American Diabetes Association, Arlington, VA 2Cedars-Sinai Medical Center, Los Angeles, CA 3McMaster University and Hamilton Health Sci- ences, Hamilton, Ontario, Canada 4Diabetes Trial Unit, Oxford Centre for Diabetes, Endocrinology and Metabolism, Oxford, U.K. 5Lunenfeld-Tanenbaum Research Institute, In December 2008, the U.S. Food and Drug Administration issued guidance to Mount Sinai Hospital, University of Toronto, Tor- the pharmaceutical industry setting new expectations for the development of onto, Ontario, Canada 6 antidiabetes drugs for . This guidance expanded the scope and cost Diabetes Research Institute, University of Mi- ami Leonard M. Miller School of Medicine, Mi- of research necessary for approval of such drugs by mandating long-term ami, FL cardiovascular outcomes trials (CVOTs) for safety. Since 2008, 9 CVOTs have been 7Duke University Medical Center, Durham, NC reported, 13 are under way, and 4 have been terminated. Reassuringly, each of 8University of North Carolina School of Medicine, the completed trials demonstrated the noninferiority of their respective drugs to Chapel Hill, NC 9Yale School of Medicine and Yale New Haven placebo for their primary cardiovascular (CV) composite end point. Notably, four Hospital, New Haven, CT additionally provided evidence of CV benefit in the form of significant decreases in 10Keenan Research Centre for Biomedical Science, the primary CV composite end point, two suggested reductions in CV death, and Li Ka Shing Knowledge Institute, St. Michael’sHos- pital, and Departments of Medicine and Nutritional three suggested reductions in all-cause mortality. Although these trials have Sciences, University of Toronto, Toronto, Ontario, yielded much valuable information, whether that information justifies the Canada investment of time and resources is controversial. In June 2016, a Diabetes Care 11Diabetes Unit, Department of Internal Medi- Editors’ Expert Forum convened to review the processes and challenges of CVOTs, cine, Hadassah Hebrew University Hospital, Jer- fi usalem, Israel discuss the bene ts and limitations of their current designs, and weigh the merits of 12Dallas Diabetes Research Center at Medical modifications that might improve the efficiency and clinical value of future trials. City and The University of Texas Southwestern Discussion and analysis continued with the CVOT trial results released in June 2017 at Medical Center, Dallas, TX 13 the American Diabetes Association’s Scientific Sessions and in September 2017 at the Oregon Health & Science University, Portland, OR European Association for the Study of Diabetes scientificmeeting.Thisarticlesum- fi Corresponding author: William T. Cefalu, wcefalu@ marizes the discussion and ndings to date. diabetes.org. Received 24 October 2017 and accepted 24 October 2017. In December 2008, the U.S. Food and Drug Administration (FDA) issued guidance to the This article contains Supplementary Data online pharmaceutical industry setting out new expectations for the development of antidia- at http://care.diabetesjournals.org/lookup/ betes drugs for type 2 diabetes (1). This guidance focused specifically on cardiovascular suppl/doi:10.2337/dci17-0057/-/DC1. (CV) safety, largely in recognition of the excess burden of cardiovascular disease (CVD) This article is featured in a podcast available at in type 2 diabetes (2). The FDA was responding to prevailing concerns about the http://www.diabetesjournals.org/content/ potential for increased CVD risk associated with certain antidiabetes drugs, notably diabetes-core-update-podcasts. the (TZD) (3,4). The guidancedand subsequent similar © 2017 by the American Diabetes Association. d Readers may use this article as long as the work requirements from the European Medicines Agency (5) effectively expanded the is properly cited, the use is educational and not scope and cost of research necessary to secure approval of new antidiabetes drugs for profit, and the work is not altered. More infor- by mandating long-term safety trials. mation is available at http://www.diabetesjournals In the 9 years since the guidance was issued, 9 long-term, prospective CV outcomes .org/content/license. trials (CVOTs) have been completed, 13 are under way, and 4 were started but termi- See accompanying articles, pp. 11, nated early. Collectively, these studies have involved more than 190,000 participants. 136, 143, 150, and 156. care.diabetesjournals.org Cefalu and Associates 15

The nine completed trials examined three article summarizes the proceedings of the Less well understood was the CV safety dipeptidyl peptidase 4 (DPP-4) inhibitors Expert Forum and CVOT findings to date. profile of available antidiabetes drugs. Be- (6–8), four -like peptide 1 (GLP- fore the FDA issued its 2008 guidance, ap- 1) receptor agonists (9–12), and two PRE-GUIDANCE RESEARCH AND proval decisions for such sodium–glucose cotransporter 2 (SGLT2) CONTROVERSY focused largely on short-term glycemic ef- fi inhibitors (13,14). Each demonstrated the Recent data indicate that, because of ad- cacy (i.e., A1C reduction), along with noninferiority of their respective drugs to vances in research and clinical care, morbid- safety data from 6- to 12-month phase placebo in their major adverse cardiac ity and mortality have decreased significantly 2 and 3 randomized clinical trials. These event (MACE) primary composite end in both type 1 and type 2 diabetes. However, trials typically enrolled younger cohorts point. Four additionally provided evi- individuals with diabetes still have greater CV with relatively recent onset of diabetes dence of CV benefit in the form of signif- risk than those without diabetes (15). Thus, and low CV risk. Indeed, individuals with icant decreases in the MACE primary additional strategies to reduce this risk established CVD were usually excluded. composite end point (10,11,13,14), two continue to be evaluated. Thus, these studies had low CV event found reductions in CV death (10,13), Well before 2008, diabetes was known rates, and the events that did occur were fi and three showed reductions in all-cause to more than double the risk for CV events neither prespeci ed as end points nor in- mortality (10,12,14)dalthough the sta- and was considered by some a coronary dependently adjudicated. As a result, esti- tistical robustness of findings for these artery disease equivalent (16). Although mates of the CV safety of these drugs were inconsistent and unreliable (31,32). secondary end points in some cases may several major clinical trials of more versus Concerns about this gap in assessment be controversial. less stringent glycemic control initially of CV safety grew after two highly contro- The completed CVOTs have provided failed to demonstrate that intensive glu- versial meta-analyses of CV risk (3,33) were much valuable information, and because cose lowering significantly reduces CV risk published in the mid-2000sdthe first on numerous antidiabetes drugs have been (17–23), meta-analyses indicated a mod- , an investigational dual per- developed and tested in the past decade, estly reduced risk of nonfatal myocardial oxisome proliferator–activated receptor it appears that the FDA mandate has not infarction (MI) (24,25). In addition, long- (PPAR)-a and -g agonist, and the second discouraged pharmaceutical companies term noninterventional follow-up of the on the FDA-approved TZD (and PPAR-g from pursuing approval for potentially DCCT (Diabetes Control and Complica- tions Trial) (26,27) in type 1 diabetes agonist) rosiglitazone. The muraglitazar successful drugs. Still, questions remain and the UKPDS (UK Prospective Diabetes analysis showed more than twice the in- regarding whether the information ob- Study) (28) and VADT (Veterans Affairs cidence of CV complications, including tained through CVOTs designed according Diabetes Trial) (29) in type 2 diabetes sug- congestive heart failure (HF) and death, to the FDA mandate justifies the time and gested an eventual “legacy” CV benefitof with the drug compared with standard resources needed, especially in light of initially tighter glycemic control (Table 1) therapy (33). Although this drug had re- the neutral results of many of the studies. (17–22,26–30), but one that was not as ceived a vote for approval from the FDA’s ’ In June 2016, a Diabetes Care Editors Expert apparent as the effect on microvascular Endocrinologic and Metabolic Drugs Ad- Forum convened and began to review the complications. However, increased CV visory Committee (34), these findings ef- processes and challenges of CVOTs, dis- mortality in patients assigned to a more fectively ended its development. The cuss the benefits and drawbacks of their intensive glucose-lowering strategy in the meta-analysis of rosiglitazone, which had current designs, and weigh the merits of ACCORD (Action to Control Cardiovascu- already been marketed for a decade, potential modifications that might im- lar Risk in Diabetes) trial (20,30) and no suggested a statistically significant 43% prove the efficiency and clinical value of reduction in CV mortality risk in the VADT increased risk of MI and a statistically future trials. This process continued (22,29) tempered enthusiasm for strin- nonsignificant 64% increased risk of CV through September 2017 to include addi- gent glucose lowering, particularly in death versus comparators (3). In the tional CVOT data reported at that time. This older patients at high CV risk. wake of these publications and the heated

Table 1—Early major trials evaluating the effects of intensive glycemic control of diabetes Study Diabetes type CV composite MI CV mortality All-cause mortality DCCT/EDIC (17,26,27) Type 1 ↔↓dd d d ↔↓ UKPDS Type 2 Main randomization (SU or vs. conventional therapy) (18,28) dd↔↓ dd ↔↓ Additional randomization of overweight patients ( vs. SU vs. conventional therapy) (19,28) dd↓* ↓* dd ↓* ↓* ACCORD (20,30) Type 2 ↔↔↓↔↑↑ ↑ ↔ ADVANCE (21) Type 2 ↔† ↔↔ ↔ VADT (22,29) Type 2 ↔↓↔↔↔↔↔ ↔ Left columns show initial results; right columns show long-term follow-up. ↔, Neutral effect; ↓, decrease; ↑, increase; d, not assessed/reported; ADVANCE, Action in Diabetes and Vascular Disease: Preterax and Diamicron MR Controlled Evaluation; SU, . Adapted from Bergenstal et al. (97). *Metformin group only. †A decrease was reported in a combined CV/microvascular composite but was found to be mostly attributable to nephropathy. 16 Cardiovascular Outcomes Trials in Type 2 Diabetes Diabetes Care Volume 41, January 2018

controversy they ignited, the FDA issued It was suggested that separate CVOTs sufficient events in a timely manner. its CV safety guidance. would be necessary only if meta-analyses Some had active run-in periods to en- After a subsequent study specifically could not rule out an unacceptable level hance adherence to the study drug. Their designed to test the CV safety of rosiglitazone of risk. However, the number of CV events participants have a long duration of dia- did not confirm the earlier findings (35,36), accrued during a typical phase 2 or 3 de- betes (mean 7.1–16.4 years), with base- the FDA lifted its prescription restrictions velopment program is usually insufficient line mean A1C ranging from 7.2 to 8.7%. for rosiglitazone (37). Even so, use of ro- to provide high statistical confidence. Thus, The characteristics of the completed siglitazone remains negligible. Trials of as noted in a review by Regier et al. (42), CVOTs and their primary findings are another TZD, , have shown every new antidiabetes drug approved summarized in Table 3 (6–14,59,60). On- CV benefit in high-risk patients. The PRO- since 2008 has had a dedicated CVOT to going CVOTs are summarized in Supplementary active (Prospective Pioglitazone Clinical evaluate whether the upper limit of a Table1(44–56), and Supplementary Table 2 Trial in Macrovascular Events) study (38) two-sided 95% CI of the estimated hazard summarizes additional trials that assessed found that pioglitazone reduced the sec- ratio (HR) is ,1.8, the requirement nec- CV outcomes but were not initiated as a ondary composite end point of all-cause essary to gain initial approval. Although direct result of the 2008 FDA guidance mortality, nonfatal MI, and stroke in pa- the FDA’s guidance specifically excluded (39,57,58,61). tients with type 2 diabetes who were at insulin from its CV safety evaluation, not- Most of the CVOTs were designed as high risk for macrovascular events. The ing that its necessity as a lifesaving therapy phase 4 event-driven trials requiring .600 more recent IRIS (Insulin Resistance Inter- for type 1 diabetes makes such evaluation primary end point events to rule out with vention After Stroke) trial (39,40) found impractical, some newer insulin formula- statistical confidence an HR with an upper that, in insulin-resistant patients (many tions have been studied in CVOT trials confidence limit $1.3 for a 3-point MACE with prediabetes but none with diabetes) based on findings from meta-analysis of (CV death, nonfatal MI, and nonfatal with previous ischemic stroke or transient phase 2 and 3 studies (43). stroke). In three trials (TECOS [Trial Eval- ischemic attack, treatment with this drug uating Cardiovascular Outcomes with significantly reduced fatal and nonfatal SUMMARY OF CVOTS TO DATE ] [8], ELIXA [Evaluation of MI and stroke. Trial Designs in Acute Coronary Syndrome] Figure 1 displays a timeline of completed [9]), and FREEDOM-CVO [A Study to Eval- REVIEW OF THE GUIDANCE and ongoing diabetes CVOTs, as well as uate Cardiovascular Outcomes in Patients The FDA, following its usual process, cast several related trials that assessed CV With Type 2 Diabetes Treated With ITCA its expectations as recommendations outcomes but were not initiated as a di- 650] [53]) a 4-point MACE, adding hos- rather than requirements, but also noted rect result of the 2008 FDA guidance pitalization for unstable angina to the that they were “for immediate implemen- (6–14,39,44–58). These trials were mainly components of the 3-point MACE, is the tation to ensure that relevant issues re- designed to rule out unacceptable CV risk, primary end point. In the ACE ( lated to minimizing cardiovascular risk are but some were powered to estimate su- Cardiovascular Evaluation) trial (58) of the considered in ongoing drug development periority after noninferiority was dem- a-glucosidase inhibitor acarbose, conducted programs” (1). The guidance outlined a onstrated. They typically have studied in China in patients with coronary heart detailed process for evaluating the CV populations in which some or all patients disease and impaired glucose tolerance safety of type 2 diabetes drugs in devel- had advanced atherosclerotic CV risk or (IGT), the primary end point is a 5-point opment (Table 2) (1,41). established CVD to ensure accrual of MACE, adding hospitalization for unstable

Table 2—Summary of the 2008 FDA guidance on CVOTs (1) To adequately evaluate the CV safety of type 2 diabetes drugs in development, future development programs should include:

c Phase 2 and 3 trials that include patients at higher risk for CV events, are of sufficient size and duration to enable enough CV events to allow for a meaningful evaluation of CV risk, and are designed to facilitate later meta-analysis; the CV events should include CV mortality, MI, and stroke and can also include hospitalization for ACS, urgent revascularization procedures, and other end points such as HF hospitalization* c Independent adjudication of CV events c Meta-analysis of the phase 2 and 3 trials to be conducted at the end of the research program, following a protocol developed in advance that prespecifies the end points to be assessed and the statistical methods to be employed c Analysis of premarketing data comparing the CV events occurring with the agent to those occurring with the control group and demonstrating that the upper limit of a two-sided 95% CI of the estimated risk ratio is ,1.8; if this cannot be done through the meta-analysis described above, it should be accomplished in a separate, large CV safety trial c For agents whose 95% CI upper limit falls between 1.3 and 1.8 in premarketing analysis, completion of a postmarketing trial or continuation of a premarketing trial after approval may be needed to conclusively show that the upper limit of the two-sided 95% CI is ,1.3 with a “reassuring” point estimate of overall CV risk *The FDA has gained valuable experience from CVOTs and CV risk assessments intended to meet the December 2008 guidance. The FDA has acknowledged that allowing MACE+ (a composite of CV death, nonfatal MI, or nonfatal stroke plus the less-specific end point of unstable angina requiring either hospitalization or revascularization) in the premarket CV risk assessment to exclude a 95% CI upper limit $1.8 is acceptable because it allows for a more reasonable sample size for evaluation. Provided no countervailing safety finding is observed, MACE+ in the premarket risk assessment strikes an appropriate balance of safety assessment without an undue burden to companies bringing new therapies to market. However, to provide longer-term, more reassuring CV safety data in the postmarketing setting, the exclusion of a 95% CI upper limit $1.3 should rely on the MACE composite (CV death, nonfatal MI, and nonfatal stroke). Furthermore, these data should come from a dedicated CV trial and not from the meta-analysis of multiple phase 2 or 3 trials (41). care.diabetesjournals.org Cefalu and Associates 17

Figure 1—Completed and ongoing CVOTs (6–14,39,44–58). 3-P, 3-point; 4-P, 4-point; 5-P, 5-point. DECLARE-TIMI 58, Multicenter Trial to Evaluate the Effect of Dapagliflozin on the Incidence of Cardiovascular Events; ESRD, end-stage renal disease; HARMONY Outcomes, Effect of , When Added to Standard Blood Glucose Lowering Therapies, on Major Cardiovascular Events in Subjects With Type 2 Diabetes Mellitus; PIONEER 6, A Trial Investigating the Cardiovascular Safety of Oral in Subjects With Type 2 Diabetes; REWIND, Researching Cardiovascular Events With a Weekly in Diabetes; VERTIS CV, Cardiovascular Outcomes Following Ertugliflozin Treatment in Type 2 Diabetes Mellitus Participants With Vascular Disease. angina or HF to the components of the baseline patient characteristics, and de- In the ELIXA trial of the GLP-1 receptor 3-point MACE. Additionally, most analysis signs make it difficult to compare results agonist lixisenatide (9), there were statis- plans permitted testing for superiority to among these trials. tically significant reductions in A1C at placebo if the noninferiority HR threshold study end (0.3%), systolic blood pressure of ,1.3 were achieved, providing an op- Effects on Cardiometabolic Risk (0.8 mmHg), and weight (0.7 kg), with a portunity to show potential CV benefit. Factors slightly increased heart rate (0.4 bpm). In All of the CVOTs in populations with The impact on cardiometabolic risk fac- the LEADER ( Effect and Action diabetes were designed to promote “gly- tors with DPP-4 inhibitors was minimal in Diabetes: Evaluation of Cardiovascular cemic equipoise” between the treatment except for small reductions in A1C of Outcome Results) trial (10), treatment groups to minimize the potentially con- 0.3% at study end (6–8). with liraglutide was associated with a founding effect of differences in glycemic The EMPA-REG OUTCOME (BI 10773 small but statistically significant increase control. Accordingly, treatment intensifi- [Empagliflozin] Cardiovascular Outcome Event in heart rate (3.0 bpm), with reductions in cation with other oral antidiabetes drugs Trial in Type 2 Diabetes Mellitus Patients) mean updated A1C (0.4%), systolic blood or insulin was more prevalent in the con- trial of the SGLT2 inhibitor empagliflozin pressure (1.2 mmHg), and weight (2.3 kg) trol groups. Even so, modest between- showed a small but statistically significant and a slight increase in diastolic blood group differences in A1C changes were 0.3% A1C reduction at study end, simi- pressure (0.6 mmHg). Similar results observed at the end of most trials, with lar reductions in systolic (4–5mmHg) were found in SUSTAIN-6 (Trial to Evalu- higher values in the placebo arms. and diastolic (2 mmHg) blood pressure ate Cardiovascular and Other Long-term All primary results were analyzed in the and weight (;2 kg), and minimal in- Outcomes With Semaglutide in Subjects intention-to-treat (ITT) (as randomized) creases in LDL and HDL cholesterol, with With Type 2 Diabetes) (11) with 0.5 and or modified ITT (as treated with at least no changes observed in heart rate (13). In 1.0 mg doses of semaglutide; heart rate one dose) population, with evaluations in the CANVAS (Canagliflozin Cardiovascu- increased by 2.0 and 2.5 bpm, respec- the on-treatment or per-protocol popula- lar Assessment Study) Program (14), tively, with robust reductions in A1C at tions reported as sensitivity analyses. there were statistically significant reduc- study end (0.7 and 1.0%, respectively), sys- Most study patients were receiving tions in mean A1C updated over time tolic blood pressure (1.3 and 2.6 mmHg, near-optimal CV risk management at (0.58%), systolic (3.93 mmHg) and diastolic respectively), and weight (2.9 and 4.3 kg, baseline, as shown by a high proportion (1.39 mmHg) blood pressure, and weight respectively) and no difference in diastolic of patients receiving antihypertensive, (1.6 kg) and increases in LDL and HDL cho- blood pressure. Finally, in the EXSCEL lipid-lowering, and antiplatelet medica- lesterol (4.68 and 2.05 mg/dL, respec- ( Study of Cardiovascular Event tions. Differences in study populations, tively) with canagliflozin treatment (14). Lowering) trial (12) with once-weekly 8Crivsua ucmsTil nTp Diabetes 2 Type in Trials Outcomes Cardiovascular 18

Table 3—CVOTs completed after issuance of the FDA 2008 guidance SGLT2 inhibitors DPP-4 inhibitors GLP-1 receptor agonists EMPA-REG CANVAS SAVOR-TIMI 53 (6) EXAMINE (7,59) TECOS (8) ELIXA (9) LEADER (10) SUSTAIN-6 (11)* EXSCEL (12) OUTCOME (13,60) Program (14) (n = 16,492) (n = 5,380) (n = 14,671) (n = 6,068) (n =9,340) (n = 3,297) (n = 14,752) (n =7,020) (n = 10,142) Intervention / / Sitagliptin/ Lixisenatide/ Liraglutide/ Semaglutide/ Exenatide QW/ Empagliflozin/ Canagliflozin/ placebo placebo placebo placebo placebo placebo placebo placebo placebo Main inclusion Type 2 diabetes Type 2 diabetes Type 2 diabetes Type 2 diabetes Type 2 diabetes Type 2 diabetes Type 2 diabetes Type 2 diabetes Type 2 diabetes criteria and history of and ACS within and preexisting and an acute and preexisting and preexisting with or and preexisting and preexisting or multiple risk 15–90 days CVD coronary CVD, kidney CVD, HF, or CKD without CVD, with CVD at factors for CVD before event within disease, or HF at $50 years preexisting BMI #45 kg/m2 $30 years randomization 180 days at $50 years of of age or CVD and eGFR $ of age or before age or $1CV $1CVrisk 30 mL/min/ $2CVrisk screening risk factor factor 1.73 m2 factors at at $60 years at $60 years $50 years of age of age of age A1C inclusion criterion (%) $6.5 6.5–11.0 6.5–8.0 5.5–11.0 $7.0 $7.0 6.5–10.0 7.0–10.0 7.0–10.5 Age (years)† 65.1 61.0 65.4 60.3 64.3 64.6 62 63.1 63.3 BMI (kg/m2)† 31.1 28.7 30.2 30.2 32.5 32.8 31.8 30.7 32 Diabetes duration (years)† 10.3 7.1 11.6 9.3 12.8 13.9 12 57% .10 13.5 Events planned/ observed (n) 1,040/1,222 650/621 1,300/1,690 844/805 $611/1,302 $122/254 1,360/1,744 691/772 688/1,011 Median follow-up (years) 2.1 1.5 3.0 2.1 3.8 2.1 3.2 3.1 2.4 Statin use (%) 78 91 80 93 72 73 74 77 75 Prior CVD/CHF (%) 78/13 100/28 74/18 100/22 81/18 60/24 73.1/16.2 99/10 65.6/14.4 A1C/A1C change (%)‡ 8.0/–0.3 8.0/–0.3 7.2/–0.3 7.7/–0.3 8.7/–0.4 8.7/–0.7 or –1.0§ 8.0/–0.53 8.1/–0.3| 8.2/–0.58 Year started/reported 2010/2013 2009/2013 2008/2015 2010/2015 2010/2016 2013/2016 2010/2017 2010/2015 2009/2017 Primary outcome¶ 3-point MACE 3-point MACE 4-point MACE 4-point MACE 3-point MACE 3-point MACE 3-point MACE 3-point MACE 3-point MACE 1.00 (0.89–1.12) 0.96 (95% UL #1.16) 0.98 (0.89–1.08) 1.02 (0.89–1.17) 0.87 (0.78–0.97) 0.74 (0.58–0.95) 0.91 (0.83–1.00) 0.86 (0.74–0.99) 0.86 (0.75–0.97)# Key secondary Expanded MACE 4-point MACE 3-point MACE Expanded MACE Expanded MACE Expanded MACE Individual 4-point MACE All-cause and outcome¶ 1.02 (0.94–1.11) 0.95 (95% UL #1.14) 0.99 (0.89–1.10) 1.00 (0.90–1.11) 0.88 (0.81–0.96) 0.74 (0.62–0.89)components 0.89 (0.78–1.01) CV mortality ibtsCare Diabetes of MACE (see below) (see below) CV death¶ 1.03 (0.87–1.22) 0.85 (0.66–1.10) 1.03 (0.89–1.19) 0.98 (0.78–1.22) 0.78 (0.66–0.93) 0.98 (0.65–1.48) 0.88 (0.76–1.02) 0.62 (0.49–0.77) 0.96 (0.77–1.18)** 0.87 (0.72–1.06)# †† – – – – – – – – – MI¶ 0.95 (0.80 1.12) 1.08 (0.88 1.33) 0.95 (0.81 1.11) 1.03 (0.87 1.22) 0.86 (0.73 1.00) 0.74 (0.51 1.08) 0.97 (0.85 1.10) 0.87 (0.70 1.09) 0.89 (0.73 1.09)# 2018 January 41, Volume Stroke¶†† 1.11 (0.88–1.39) 0.91 (0.55–1.50) 0.97 (0.79–1.19) 1.12 (0.79–1.58) 0.86 (0.71–1.06) 0.61 (0.38–0.99) 0.85 (0.70–1.03) 1.18 (0.89–1.56) 0.87 (0.69–1.09)# Continued on p. 19 care.diabetesjournals.org Cefalu and Associates 19

exenatide, mean updated A1C was 0.53% lower, weight was 1.3 kg lower, and sys- tolic blood pressure was 1.6 mmHg lower, 0.77)# 0.87)# 1.07)# 1.01)** – – – – but heart rate was 2.5 bpm higher. d 20 November

= 10,142) Overall, the impact of treatment inter- CANVAS n (

Program (14) ventions on cardiometabolic risk factors 0.90 (0.76 was small, and the resulting effects of these Worsening nephropathy changes on CV outcomes remain unclear. ‡‡ 0.70) 0.60 (0.47 0.85) 0.67 (0.52 1.34) 0.82) 0.87 (0.74 SGLT2 inhibitors – – – – CV Outcomes

=7,020) DPP-4 Inhibitors n ( EMPA-REG Age and BMI were reported as means in

† Five trials enrolling 49,618 patients have 0.61 (0.53 OUTCOME (13,60)

ed. been designed to evaluate the CV safety fi of DPP-4 inhibitors. Three have reported ata from CANVAS, including before 1.13) 0.65 (0.50 1.18) 0.99 (0.74 0.97) 0.68 (0.57 outcome results (6–8), and two are ongo- – – –

d ing (44,45). One additional trial, OMNEON = 14,752)

th from renal disease in LEADER, SUSTAIN-6, and EMPA-REG (A Study to Assess Cardiovascular Out- n EXSCEL (12) (

SCEL reporting medians and EMPA-REG OUTCOME reporting as comes Following Treatment With Omari- ed exploratory adjudicated outcome in SAVOR-TIMI 53, LEADER, fi gliptin [MK-3102] in Participants With Type 2 Diabetes Mellitus [MK-3102-018]) 0.88) 1.61) 0.94 (0.78 1.44) 1.05 (0.94 1.50) 0.86 (0.77 – – – – (62), evaluating the investigational once- ed in treating hierarchy as the principal data set for analysis for superiority of fi

= 3,297) weekly omarigliptin in 4,202 patients with n ( type 2 diabetes and CVD, was terminated SUSTAIN-6 (11)*

1.8; superiority hypothesis not prespeci as a business decision (63). Although all ed integrated data (refers to pooled d $ three completed trials (6–8) met the pri- A1C change of 0.30 in EMPA-REG OUTCOME is based on pooled results for both doses (i.e., | 0.92) 0.64 (0.46 1.05) 1.11 (0.77 1.26) 0.82 (0.47 0.97) 1.05 (0.74 mary objective of excluding an HR upper – – – –

mol/L) in SAVOR-TIMI 53; as the new onset of macroalbuminuria (urine albumin creatinine limit $1.3, none were associated with m =9,340)

n fi

, A1C difference was updated over time; for SAVOR-TIMI 53, EXAMINE, ELIXA, SUSTAIN-6, and EMPA- any suggestion of CV bene t. Likewise, ( LEADER (10) rsening nephropathy was a prespeci

0.78 (0.67 at its termination after a median follow-up of 96 weeks, the OMNEON trial reported an HR of 1.00 (95% CI 0.77–1.29) for its rials except EXAMINE, ELIXA, and SUSTAIN-6, which reported for nonfatal events only. 1.23) 0.87 (0.73 2.62) 0.98 (0.76 1.13) 0.85 (0.74 6.0 mg/dL (530 – – – t four trials, with SAVOR-TIMI 58, EXAMINE, and EX

. primary 3-point MACE end point (62).

= 6,068) The SAVOR-TIMI 53 (Saxagliptin As- , the need for continuous renal-replacement therapy, or dea 2 ELIXA (9) n ( sessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus–Thrombolysis in Myocardial Infarction) trial (6) with sax- 1.20) 0.96 (0.75 1.16) 1.11 (0.47 1.14) 0.94 (0.78 – – – agliptin suggested an increased risk for incident HF (HR 1.27 [95% CI 1.07–1.51], 45 mL/min/1.73 m = 14,671) TECOS (8) # n

eath from renal causes in CANVAS. Wo P = 0.007 [6]), with some support pro- ( vided by a trend noted in the EXAMINE

ed data from CANVAS after 20 November 2012 plus CANVAS-R; prespeci (Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care) For TECOS, LEADER, EXSCEL, and the CANVAS Program 1.58) 1.00 (0.83 1.37) 0.90 (0.70 1.09) 1.01 (0.90 ‡ – – – trial with alogliptin (HR 1.19 [95% CI Reported for fatal and nonfatal events in all t

†† – ddd 0.90 1.58], P = 0.220 [7,59]), earning a = 5,380) n

( warning by the FDA, especially in patients 10 years. EXAMINE (7,59) ozin). ¶Outcomes reported as HR (95% CI) unless otherwise noted. #Nontruncat s; diabetes duration was reported as means in all bu DPP-4 inhibitors GLP-1 receptor agonists with underlying heart and kidney disease . fl (64). CAROLINA (Cardiovascular Outcome Study of Versus 1.51) 1.19 (0.90 1.32) 1.60) 0.90 (0.60 1.27) 0.88 (0.71 – – – – in Patients With Type 2 Diabetes) (44)

= 16,492) was designed to establish noninferiority n (

eath in the CANVAS Program). of linagliptin relative to the sulfonylurea 1.08 (0.88 SAVOR-TIMI 53 (6) glimepiride, thus including an active com- parator. However, because the CV safety

‡‡ of glimepiride has not been established, a placebo-controlled trial (CARMELINA [Car- Continued

— diovascular and Renal Microvascular Out- ned as doubling of creatinine level, initiation of dialysis, renal transplantation, or creatinine 300 mg/g) or a doubling of the serum creatinine level and an eGFR of fi

. come Study With Linagliptin in Patients , not assessed/reported; 95% UL, upper limit of 95% CI; CHF, congestive heart failure. *Powered to rule out an HR upper margin nephropathy¶ hospitalization¶ 1.19 (0.89 With Type 2 Diabetes Mellitus] [45]) is was de 0.24% for 10 mg and 0.36% for 25 mg of empagli REG OUTCOME, A1C difference was at study end. §A1C change of 0.66% with 0.5 mg and 1.05% with 1 mg dose of semaglutide. Table 3 d HF hospitalization¶ 1.27 (1.07 OUTCOME; and as 40% reduction in eGFR, renal-replacement therapy, or d ratio 2012 plus CANVAS-R). **Truncated integrated data setall-cause (refers mortality to and pool CV d all trials except EXAMINE, which reportedpercentage of median population with diabetes duration SUSTAIN-6, and CANVAS but not in EMPA-REG OUTCOME. Unstable angina All-cause mortality¶ 1.11 (0.96 Worsening also being conducted to establish the CV 20 Cardiovascular Outcomes Trials in Type 2 Diabetes Diabetes Care Volume 41, January 2018

and renal safety of linagliptin by including EMPA-REG OUTCOME (13) was an event- results of the primary outcome and the an enriched population with CV risk and driven trial in patients at high CV risk robustness of the CV death data, this in- evidence of renal compromise. These tri- randomly treated with 10 or 25 mg dication was approved. On 15 December als may help to answer the question of CV of empagliflozin or placebo, with a me- 2016, the American Diabetes Associa- safety with , which has been dian observation time of 3.1 years. The tion (ADA) endorsed this recommenda- lingering for nearly 50 years (65). primary hypothesis was noninferiority tion in its Standards of Medical Care in Overall, the DPP-4 inhibitors have es- for the primary 3-point MACE end Diabetesd2017 (72). Although EMPA- tablished MACE safety. point with empagliflozin (pooled doses REG OUTCOME was the first trial to demon- SGLT2 Inhibitors of 10 and 25 mg) versus placebo, with strate the CV superiority of an antidiabetes The CV and renal safety of SGLT2 in- an HR upper limit of ,1.3. The primary drug, more research is needed to elucidate hibitors is being evaluated in nine trials end point was significantly reduced with the underlying mechanisms, confirm whether enrolling 62,378 patients. Of these, twod empagliflozin treatment (HR 0.86 [95% CI the CV benefits are a class effect, and deter- EMPA-REGOUTCOME(13) and the CANVAS 0.74–0.99], P = 0.038). This reduction was mine whether empagliflozin conveys a CV Program (14)dhave been completed, driven by a significant decrease in CV benefit to patients without established CVD and each reported positive CV and renal death (HR 0.62 [95% CI 0.49–0.77], P , or diabetes. outcomes. The other seven trials are sched- 0.001), with numerical, but not statisti- The CANVAS Program was originally uled for completion within the next 2–3 cally significant, differences in nonfatal designed in two phases to establish the years (46–52). MI (favoring empagliflozin) and nonfatal CV protective effects of canagliflozin in In four SGLT2 inhibitor trials, a compos- stroke (favoring placebo). There was no high-risk patients with type 2 diabetes ite 3-point MACE is the primary end point; difference in the key secondary end point and established CVD or risk factors for in two, a composite renal outcome is the of 3-point MACE plus hospitalization for CVD (73). The first phase was included primary end point; and in three, a com- unstable angina (HR 0.89 [95% CI 0.78– in the meta-analysis presented to the posite of HF outcomes and CV death is the 1.01], P = 0.08). Significant reductions in FDA for regulatory approval (74), which primary end point in people with estab- prespecified end points of all-cause mor- ruled out an HR upper limit $1.8. How- lished HF. Although primarily renal stud- tality (32%) and HF hospitalization (35%) ever, these results were publicly disclosed ies, CANVAS-R (CANVAS-Renal) (66) (data were also observed. Superiority in terms at an advisory committee meeting in from which were included in the com- of CV death was large and clinically im- 2013 after partial (group-level) unblinding bined report on the overall CANVAS Pro- portant (38% risk reduction), statistically due to an observed dose-dependent in- gram [14]) and the ongoing CREDENCE robust (P , 0.001), based on a large num- crease in LDL cholesterol. The FDA felt (Evaluation of the Effects of Canagliflozin ber of events (n = 309), occurring early that the lipid data necessitated public dis- on Renal and Cardiovascular Outcomes in (within 3 months), consistently seen closure, and with it, the CV event data Participants With Diabetic Nephropathy) with both doses across subgroups, and were also disclosed, which had the poten- trial (46) also prospectively collected/are established across sensitivity analyses tial to compromise the postmarketing collecting adjudicated CV outcomes, as that accounted for silent MIs (which phase of the trial to fulfill the requirement required by the FDA mandate. Dapa- were excluded from event adjudication), of ruling out an HR upper limit $1.3. Be- CKD (A Study to Evaluate the Effect of nonassessable deaths, and missing data. cause of these extenuating circumstances, the Dapagliflozin on Renal Outcomes and Car- There was an increased rate of treatment- FDA accepted an integrated analysis plan, diovascular Mortality in Patients With emergent genital infection but no increase whereby data from extended follow-up of Chronic Kidney Disease) (50) will also in other adverse events, including diabetic the CANVAS first phase before and after the evaluate CV death or HF hospitalization ketoacidosis, bone fractures, venous throm- data disclosure (20 November 2012) would in addition to the primary composite re- boembolism, hypoglycemic events, amputa- be combined with new data from CANVAS-R nal outcome. Plans for a dedicated study tions, and acute kidney injury. Interestingly, (66) to address CV safety as assessed by to evaluate empagliflozin on the progres- reductions in CV outcomes appeared to be exclusion of a postmarketing risk ratio of sion of CKD were recently announced unexplained by empagliflozin’sglucose- 1.3. A sequential hypothesis testing plan (67). Although primarily HF studies, lowering effect or by its effects on blood was used, whereby a truncated integrated Dapa-HF (Study to Evaluate the Effect of pressure, weight, or uric acid, suggesting data set of pooled data from CANVAS after Dapagliflozin on the Incidence of Worsen- that other mechanisms activated by 20 November 2012 plus CANVAS-R was ing Heart Failure or Cardiovascular Death SGLT2 inhibition may contribute to this prespecified as the principal data set for in Patients With Chronic Heart Failure drug’s cardioprotective effects (68–70). analysis for superiority of all-cause mortal- [49]) and EMPEROR-Reduced (Empagli- Based on this favorable benefit-risk pro- ity and CV death (75). flozin Outcome Trial in Patients With file, the FDA approved a new indication for The primary 3-point MACE end point was Chronic Heart Failure with Reduced Ejec- empagliflozin on 2 December 2016 “to re- significantly reduced with canagliflozin tion Fraction [51]), both of which are in HF duce the risk of cardiovascular death in treatment (HR 0.86 [95% CI 0.75–0.97], patients with reduced ejection fraction, adult patients with type 2 diabetes melli- P = 0.02), thereby meeting the crite- and EMPEROR-Preserved (Empagliflozin tus and cardiovascular disease” (71), a rion for noninferiority using the 1.3 risk Outcome Trial in Patients With Chronic claim that represented a clinical break- margin (14). It is debatable, from a rigor- Heart Failure With Preserved Ejection throughintreatingpatientswithtype2 ous statistical perspective, whether an in- Fraction [52]), in HF patients with pre- diabetes. It was also a regulatory break- ference of superiority is justified because served ejection fraction, will also prospec- through because CV death was a second- it was apparently not prespecified in the tively evaluate adjudicated renal outcomes. ary outcome but, in view of the positive testing sequence. Because the all-cause care.diabetesjournals.org Cefalu and Associates 21

mortality difference was not statistically analyses. All-cause mortality, but not HF carcinoma; and serious adverse events significant in the truncated integrated hospitalization, was also significantly re- including severe hypoglycemia did not data set (HR 0.90 [95% CI 0.76–1.07], P = duced with liraglutide. An interesting ob- differ significantly between the two treat- 0.24), given the statistical rules of the pre- servation is the delayed separation of the ment groups. specified hierarchical testing strategy, all Kaplan-Meier curves (indicating time to Of interest, the treatment persistence subsequent results, including the favor- benefit) in LEADER (.12 months for CV with weekly exenatide was low, with 43% able effects on HF hospitalization, pro- death and .18 months for all-cause drug discontinuation. The pragmatic na- gression of albuminuria, and worsening deaths and HF hospitalization) (10), sug- ture of the study design, with visits every nephropathy, were deemed exploratory. gesting that the CV benefit may be con- 6 months and limited study support, may The point estimate for nonfatal or total veyed via an impact on atherosclerotic explain the low treatment adherence and stroke was favorable for canagliflozin. disease, which contrasts with the earlier persistence. It is remarkable that despite There was an increased rate of treatment- time to benefit(,3 months) in the EMPA- this limited drug exposure and a hetero- emergent genital infection as expected, REG OUTCOME trial (13). On 25 August geneous population of whom 27% had no but no increase in venous thromboem- 2017, based on the results of the LEADER history of CVD, the 3-point MACE reduc- bolism or hypoglycemic events. Diabetic trial, the FDA approved a new indication tion of 9% came so close to reaching sta- ketoacidosis was very low but slightly nu- for liraglutide “to reduce the risk of major tistical significance, with HRs of almost all merically increased with canagliflozin, and adverse cardiovascular events in adults measured parameters in the direction of there was a nearly twofold increase in risk with type 2 diabetes mellitus and estab- benefit. for lower-leg amputation (6.3 and 3.4 lished cardiovascular disease” (76). Although full results have not yet been events/1,000 patient-years for canagliflozin SUSTAIN-6 confirmed the noninferior- published, the phase 3 FREEDOM-CVO tri- and placebo, respectively) and a 26% rela- ity of once-weekly treatment with 0.5 al, which evaluated continuous delivery tive increase in risk of bone fracture with or 1 mg of the long-acting semaglutide, of exenatide and was designed for safety canagliflozin treatment. The underlying which is currently under FDA review (11). purposes to accrue a limited number of mechanisms for these findings remain un- This phase 3 trial was designed to rule CV events, reported in a press release (53) clear. There are some potential biological out an HR upper limit $1.8 and achieved that the primary objective of achieving an explanations for bone fractures, but none statistical superiority (HR 0.74 [95% CI HR upper limit ,1.8 had been met, and a for the small but significant and unexpected 0.58–0.95], P , 0.001), but superiority new drug application has been submitted increase in lower-leg amputations. Neither analysis was not prespecified. The favor- to the FDA (78). Treatment persistence bone fracture nor lower-limb amputation able effect on the 3-point MACE was ac- could not have been an issue in this trial has been documented with the other companied by a significant decrease in given the continuous delivery of exena- SGLT2 inhibitors, so additional evaluation nonfatal stroke (HR 0.61 [95% CI 0.38– tide via a subcutaneously implanted is need to clarify whether this should be 0.99], P = 0.04) and a nonsignificant de- mini-osmotic pump. However, the limited viewed as a class effect. crease in nonfatal MI (HR 0.74 [95% CI number of participants prevents any de- GLP-1 Receptor Agonists 0.51–1.08], P =0.12),withnotrendfor finitive conclusions beyond meeting the The CV safety of GLP-1 receptor agonists CV death or all-cause mortality. The risk HR requirement for regulatory approval. has been assessed in eight trials enrolling reduction for the primary outcome was The heterogeneity in CV outcomes (i.e., 60,090 patients, of which four have re- seen despite an increase in pulse rate, null effects with lixisenatide and exena- ported outcomes (9–12). A fifth trial has an effect seen in all CVOTs in this drug tide and favorable effects with liraglutide been completed but not reported (53), and class to date. On 18 October 2017, the and semaglutide) might be mainly the re- the other three are scheduled for comple- FDA’s Endocrinologic and Metabolic sult of differences in study populations, tion within the next 1–2 years (54–56). Drugs Advisory Committee voted in favor trial designs, and treatment persistence. The first reported, ELIXA (9), which was of approval for semaglutide (77). However, it could also reflect differences conducted in patients with a history of The fourth trial, EXSCEL (12), was per- in pharmacokinetic and pharmacodynamic recent acute coronary syndrome (ACS), formed in a usual-care setting among pa- properties (i.e., short-acting lixisenatide was CV neutral, confirming the noninfer- tients with type 2 diabetes with or without is a once-daily prandial drug that acts pri- iority of lixisenatide with respect to a previous CVD. It confirmed the noninfer- marily on postprandial glucose, whereas 4-point MACE but showing no beneficial iority, but not superiority, of once-weekly liraglutide and semaglutide are longer- effect on any CV outcome. treatment with 2 mg of the long-acting acting drugs that act mainly on fasting The second trial, LEADER (10), demon- extended-release exenatide (HR 0.91 glucose with a carryover effect on post- strated the CV noninferiority, as well as [95% CI 0.83–1.00], P = 0.06). The risk of prandial glucose) or other trial or drug the statistical superiority, of once-daily death from any cause was 6.9% in the differences, including in structural simi- treatment with liraglutide. The reduction exenatide group and 7.9% in the placebo larity to human GLP-1. What is clear so in 3-point MACE (HR 0.87 [95% CI 0.78– group (HR 0.86 [95% CI, 0.77–0.97]). This far is that drugs in this class, if tolerated, 0.97]) with liraglutide was driven by a sig- difference was considered only explor- are safe, and, if affordable, should be con- nificant reduction in CV death (HR 0.78 atory on the basis of the hierarchical test- sidered more often for type 2 diabetes [95% CI 0.66–0.93]) with numerical, but ing plan. The rates of CV death, fatal or management. not statistically significant, differences in nonfatal MI, fatal or nonfatal stroke, HF Other Drugs nonfatal MI and nonfatal stroke favoring hospitalization, and ACS hospitalization; PPAR Agonists. Two trials to determine the liraglutide. The P value for the 3-point the incidences of acute pancreatitis, pan- CV effects of , a dual activator MACE met superiority in all sensitivity creatic cancer, and medullary thyroid of PPAR-a and -g, were started but 22 Cardiovascular Outcomes Trials in Type 2 Diabetes Diabetes Care Volume 41, January 2018

terminated early due to safety concerns more light on the relative CV effects of to our overall understanding of the CV and futility for efficacy. The AleCardio trial sulfonylureas. effects of pharmacotherapies for type 2 (79) enrolled 7,226 patients with a recent Insulin. Although undertaken 5 years be- diabetes. ACS event. At its termination after a me- fore the 2008 FDA guidance and thus out- dian follow-up of 104 weeks, several CV side of the primary focus of this article, Benefits and Challenges of the Current and non-CV safety concerns were identi- the ORIGIN Trial (Outcome Reduction CVOT Design fied, including increased risks of HF and with an Initial Glargine Intervention) Benefits bone fractures, decreased estimated glo- (82,83), established the neutral CV effects Trials have shown that glargine and merular filtration rate (eGFR), and evi- of compared with stan- degludec, sitagliptin, alogliptin, saxagliptin, dence of increased risk of gastrointestinal dard glycemic care in 12,537 participants lixisenatide, and once-weekly exenatide bleeding, which was unanticipated. The with CV risk factors plus impaired fasting (6–9,12,57,82,83) have neutral effects ALEPREVENT trial (80) enrolled 1,999 pa- glucose, IGT, or type 2 diabetes who were on MACE outcomes, thereby supporting tients with prediabetes or type 2 diabetes followed for a median of 6.2 years. the use of these drugs when needed and established stable CVD. At its termina- More recently, DEVOTE (A Trial Com- to improve glycemic control with the goal tion after a brief treatment period (58 6 paring Cardiovascular Safety of Insulin of limiting microvascular complications 38 days), aleglitazar was associated with Degludec Versus Insulin Glargine in Pa- without increasing CV risk. Well-designed increased incidences of hypoglycemia tients With Type 2 Diabetes at High Risk and adequately powered CVOTs for other and muscular events. of Cardiovascular Events) (57), involving antidiabetes drugs have also provided valu- As previously mentioned, the IRIS trial 7,637 patients with established CVD able information beyond their primary of the TZD pioglitazone (39,40) reported a and a mean diabetes duration of 16 years, safety purpose. The evidence for CV benefit significant 24% reduction in its composite confirmed the CV safety of insulin deglu- from empagliflozin, canagliflozin, liraglutide, end point of fatal or nonfatal stroke or MI dec compared with insulin glargine (HR and possibly semaglutide versus placebo compared with placebo after 4.8 years of 0.91 [95% CI 0.78–1.06], P =0.21).This (10,11,13,14), supplemented by data follow-up in 3,876 insulin-resistant sub- study was conducted after the FDA from other studies, already has prompted jects with a recent ischemic stroke or required a preapproval degludec CVOT reconsideration of treatment guidelines transient ischemic attack but without di- based on the results of its phase 2 and (70,84) and is likely to alter clinical prac- abetes. No between-group difference in 3 meta-analysis (43). Degludec was statis- tice and reduce CV events and deaths all-cause mortality was found. tically superior with regard to hypoglyce- among people similar to those enrolled TOSCA.IT ( or Sulfo- mia risk, with a lower rate of both severe in these trials. nylureas Cardiovascular Accidents Inter- and nocturnal severe hypoglycemia (by Other insights from these trials have vention Trial) (81), a trial comparing the 40 and 53%, respectively; P , 0.001 for derived from secondary end points, which CV safety of pioglitazone to a sulfonyl- both comparisons). These findings corre- are in some cases less statistically robust urea in 3,028 patients, was terminated sponded to absolute incidence rates for but nonetheless of potential importance. on the basis of a futility analysis after a severe hypoglycemia of 4.9 and 6.6% and Notably, recent CVOTs have focused median follow-up of 57 months. A total of for severe nocturnal hypoglycemia of 1.0 attention on the pressing problem of 213 events were adjudicated and in- and 1.9% for degludec and glargine, re- HF, which affects older people with dia- cludedintheCVanalysis.Theprimarycom- spectively. Interestingly, despite the dif- betes more frequently than MI (85). posite end point (nonfatal MI, nonfatal ferences in severe hypoglycemia, no Empagliflozin caused an early, large, and stroke, all-cause death, and urgent coro- differences in CV mortality were found. somewhat unexpected reduction in hos- nary revascularization) showed no signif- a-Glucosidase Inhibitors. The ACE trial of pitalization for or death from HF (HR icant difference between pioglitazone acarbose (58) in 6,522 Chinese patients 0.61 [95% CI 0.47–0.79], P , 0.001) (86). and sulfonylurea treatment (HR 0.96 with coronary heart disease and IGT Canagliflozin also had favorable effects on [95% CI 0.74–1.26], P = 0.790). Pioglita- showed no relative risk reduction for the HF, although perhaps of lesser magnitude zone was associated with a reduced risk primary outcome of 5-point MACE (3-point (14). Less certain, but also of interest, is of severe hypoglycemia (,0.2 vs. 2.0%, MACE plus hospitalization for unstable an- the incidental observation of increased P , 0.0001) and moderate hypoglycemia gina or HF) (HR 0.98 [95% CI 0.86–1.11], HF with saxagliptin and a consistent non- (10 vs. 32%, P , 0.001). Despite the fact P = 0.73) or for any secondary end points significant trend with alogliptin, despite that pioglitazone showed evidence of CV compared with placebo. Although acar- no effect on other CV end points (6,59). benefit in both the PROactive (38) and bose treatment did not reduce CV risk, If this harm is confirmed, it will draw at- IRIS (39,40) trials, this study failed to rep- a modest reduction in incident type 2 tention to previously unsuspected effects licate the treatment benefit, albeit in a diabetes was observed (HR 0.82 [95% of and differences among DPP-4 inhibitors, lower-risk population. It should be noted CI 0.71–0.94], P = 0.005) over a median which block cleavage of many circulating that TOSCA.IT was ultimately underpow- follow-up of 5 years with a dose of 50 mg peptides and thus may have various down- ered, with a substantial percentage of pa- three times daily. stream effects (either benign or harmful) tients not taking their assigned medica- Although ORIGIN, IRIS, DEVOTE, and other than reducing clearance and increas- tion because of concerns that emerged ACE were not initiated as a direct result ing blood levels of GLP-1. during the trial about a possible link of the 2008 FDA guidance and OMNEON, Another clinically important benefitof between pioglitazone and bladder can- AleCardio, ALEPREVENT, and TOSCA. these trials has been assessment of the cer. A longer and larger active-controlled IT were terminated early, these trials none- effect of the tested drugs on kidney dis- study (e.g., CAROLINA [44]) should shed theless contributed important insights ease, itself a CV risk factor. For example, care.diabetesjournals.org Cefalu and Associates 23

empagliflozin (HR 0.61 [95% CI 0.53– They must be able to provide informed Experience from the UKPDS and DCCT/ 0.70], P , 0.001) (60), canagliflozin (HR consent, agree to attend the study site EDIC (Epidemiology of Diabetes Interven- 0.60 [95% CI, 0.47–0.77], P , 0.0001) for an extended period of time, and be tions and Complications) cohorts, which (14), liraglutide (HR 0.78 [95% CI 0.67– physically able to do so. To answer ques- were treated early in the course of diabe- 0.92], P = 0.003) (10), and semaglutide tions of safety and efficacy, they must be tes, shows that important medical compli- (HR 0.64 [95% CI 0.46–0.88], P = 0.005) at high risk to experience the outcome in cations of diabetes, including blindness, (11) all reduced progression of renal dis- question (in this case, a CV event or death) renal failure, amputation, CV death, and easemorethanwouldbeexpectedfrom in order to provide an adequate number of others, do not usually appear in the first the improvements in glucose and blood events in an acceptable period of time. 5 years. Rather, each has a characteristic pressure provided by these drugs. Impor- They cannot be entirely representative of and much longer development timeline tantly, the renal benefits of the SGLT2 in- the general population, and therefore the (26–28). In the case of CVD, significant hibitors extended to actual deterioration of findings of such trials can be extrapolated numbers of CV events in the UKPDS and renal function as denoted by a doubling of to a wider population only with consider- DCCT/EDIC did not accumulate until af- serum creatinine or major reduction in able caution. As an example, the EMPA- ter .10 years of follow-up. In part be- eGFR, whereas those of the GLP-1 receptor REG OUTCOME trial provides reasonable cause of the slow accrual of events, the agonists appear to be mainly on the pro- confidence that people with type 2 diabe- apparent CV benefits of early randomiza- gression to macroalbuminuria. However, tes and a history of previous CVD are likely tion to intensive glycemic control were semaglutide appears to have a deleterious to benefitfromempagliflozin therapy in not observed until long after initiation effect of diabetic retinopathy, possibly re- the same manner as did the patients in of treatment. The CV benefits likewise oc- lated to rapid and dramatic improvement in the trial itself (13). However, there is less curred long after the change in A1C dif- glycemic control in patients with preexisting certainty that empagliflozin will similarly ference disappeared after cessation of retinopathy (11). A lesser trend for ad- improve outcomes for individuals with di- randomized treatment, leading to the verse eye outcomes with liraglutide re- abetes and no history of CVD or for those concept of a “legacy effect,” for which quires further assessment (10). with CVD but without diabetes. In the thereisnoprovenexplanation.These Together, these findings have ener- CANVAS Program, one-third of partici- long-term benefits were shown through gized the medical community. Many CV pants had only CV risk factors and no evi- observational and possibly incomplete experts appear to have revised their pre- dence of CVD. Perhaps the less impressive data within which residual confounding vious skepticism about the potential for but still positive outcomes with canagliflozin cannot be excluded, so they cannot con- CV benefits from diabetes-specificthera- in this research program may be related fidently be attributed only to glycemic pies. Diabetes researchers are exploring to this population subset that might control. Nevertheless, a mechanism plau- mechanisms that may explain the clinical require longer drug exposure in a longer- sibly underlying the legacy effect is a re- effects first noted in these trials. The find- term trial (14). Notably, a subgroup anal- duction of glucose-related changes of ings with empagliflozin and canagliflozin, ysis of the LEADER trial suggested the vascular tissue structure, occurring during which alter renal clearance of sodium as hypothesis that liraglutide effectively intensive glycemic therapy but persisting well as glucose and thus cause hemody- lowered CV risk only in people with estab- thereafter, that may have slowed the early namic changes, have led to reconsidera- lished CVD (HR 0.83 [95% CI 0.74–0.93] progression of atherosclerosis and led to a tion of abnormalities in salt and water compared with HR 1.20 [95% CI 0.86– much later reduction in overt CV events. balance that are intrinsic to diabetes, an 1.67] in those without CVD, with a signif- Whether any nonglycemic effect of treat- area long inactive in physiological re- icant unadjusted interaction P value of ments in these trials contributed to this search (87). Another provocative pro- 0.04) (10). Due to the large number of benefit cannot be determined. posal is that injured myocardial and statistical tests that were performed for This view is consistent with the findings renal tissues may benefit from access to subgroups, the possibility that this was a of other large trials of short-term glucose- the energy supplied by modestly higher chance observation cannot be excluded. lowering strategies using similar treatments circulating ketone levels during treatment Nonetheless, the FDA-approved indica- that were applied later in the natural history with SGLT2 inhibitors (68,69). These hy- tion for liraglutide for CV risk reduction of type 2 diabetes. In ACCORD (20,88) and potheses open new areas of investigation was restricted only to those with estab- VADT (22,29), there were only modest re- for both clinical and basic research. lished CVD. ductions in some secondary CV end points Challenges In addition, the present CVOT design in patients with a long duration of diabetes Aside from obvious questions about com- has serious limitations related to the and high CV risk, but in ACCORD, there peting priorities for allocation of financial, natural histories of diabetes and CVD. was an unexplained early increase in mor- clinical, and professional resources and, in Because the concerns of regulatory tality. Specifically, very intensive glucose- some cases, delays in regulatory approval, agencies, pharmaceutical companies, lowering treatment goals achieved mainly the currently advised structure for CVOTs health systems, and in many cases, indi- with the use of metformin, sulfonylureas, has several practical limitations (Table 4). viduals with diabetes have relatively TZDs, and intensive insulin regimens in these The most serious of these is the lack of short timelines, current CVOTs address trials were unable to diminish the risk of generalizability of findings to the entire only short-term outcomesdnot those events in patients who already had advanced population of people with diabetes. occurring .5 years after the onset of CV pathology. As currently performed, randomized treatment. However, several lines of evi- The failure of targeting very intensive controlled trials (RCTs) require participants dence argue for the importance of longer- glucose lowering in ACCORD and VADT to have certain inclusion characteristics. term follow-up. highlights the benefits of treatment with 24 Cardiovascular Outcomes Trials in Type 2 Diabetes Diabetes Care Volume 41, January 2018

Table 4—Limitations of current CVOT structure and opportunities for improvement Current limitations

Lack of generalizability Current CVOTs include participants who are at high risk for a CV event or death and thus are not representative of the larger population. Short timeline for assessing potential benefits CV benefit may not become apparent until long after initiation of treatment. Current CVOTs do not assess outcomes occurring .5 years after the onset of treatment. Short timeline for assessing potential harm CVOTs lasting ,5 years are not likely to detect risks that may become apparent only after years of treatment. This may be especially concerning for agents with complex mechanisms of action. Placebo-controlled design Nearly all CVOTs to date have tested one drug against placebo, with both groups attempting to attain comparable glycemic control using regimens that also include other medications. Problems with this design include 1) comparable glycemic control generally has not been achieved and the lower A1Cs found with the study drug may contribute to CV benefits, 2) some drugs used in the comparison groups may themselves have an adverse effect on CV events, and 3) the CV benefits found with four agents to date make the future use of placebo ethically challenging. Opportunities for improvement Lower-risk, more diverse populations Primary intervention trials in lower-risk populations could determine whether diabetes medications offer CV protection for those who do not yet have CVD. This would require larger and/or longer studies but would yield valuable information with regard to CVD prevention. Longer-term follow-up Trial designs that prespecify longer-term follow-up could better identify longer-term safety issues and late beneficial effects, produce better cost-effectiveness data, and improve understanding of changing treatment requirements over time. Such a design would require new consent procedures to permit lifelong follow-up, strategies to increase therapy adherence and persistence, expanded use of EMRs, and innovative statistical approaches to permit serial reporting of key clinical outcomes over time. Active comparators Using an active comparator instead of placebo could address the drawbacks of placebo- controlled trials but will require sufficient knowledge of the CV impact of the comparator to avoid confounding interpretation of the results. Although challenging, this may become feasible as understanding of the CV safety of newer agents increases. Across-trials consistency in enrollment criteria and the capturing of baseline patient characteristics will facilitate such efforts. Innovative designs Adoptionof factorialor adaptivedesigns, superioritytrials, trials embedded within health care systems or networks, and/or employment of “big data” to dissect the effects of new diabetes medications may provide practical opportunities for further investigation. Standardized definitions Standard definitions of important safety and microvascular outcomes would facilitate better comparisons among agents. Collaborative efforts should be made to standardize definitions of high-priority safety and microvascular outcomes, akin to those established for CV outcomes. Modification of end points and analyses Incorporating weighted composite end points that include estimation of the severity of events, as well as multiple events in the same patient, may yield more nuanced findings in future studies. The design of trials for new agents should be informed by data from previousCVOTswithinthesamedrugclass.Importantsecondaryoutcomeswithrobust statistical findings that are biologically plausible and supportable by external evidence should be independently considered even if primary composite outcomes are not achieved. Such approaches would require buy-in from regulatory bodies and mechanismstoensureequityfordevelopersoffirst-in-classinterventionsand/orthose who willingly adopt more complex trial designs. How best to incorporate predefined safety concerns into primary analyses should also be considered. Establishment of biorepositories Future trials should obtain informed consent to store participants’ biological samples in case unexpected results warrant further investigation. Such biorepositories could increase opportunities to investigate various mechanisms contributing to CV events and key subgroups and will become increasingly important for subsequent biomarker, gut microbiota, and genomic analyses to facilitate precision medicine opportunities. Enhanced efficiency and cost-sharing options Conducting a CVOT for each new diabetes drug is cumbersome and expensive. Strategies to enhancetrial efficiency, such as collecting CV outcomes data from trials designed for other purposes, should be strongly considered, as should new models for cost-sharing among pharmaceutical, governmental, and other organizations. Involvement of patients and advocacy organizations Involving patients and their advocates in designing future trials will help to ensure that patients’ views and wishes are taken into account and that patient-related outcome measures are fully integrated. Such efforts would likely increase patient buy-in and help to minimize discontinuation, improve treatment adherence and persistence, and avoid missing data. care.diabetesjournals.org Cefalu and Associates 25

empagliflozin, canagliflozin, liraglutide, LEADER (53 over 3.8 years), and SUSTAIN- beneficial effect of the intervention stud- andsemaglutideintheirrespectiveCVOTs. 6 (44 over 2 years) were similar to those ied because the CVD is too advanced, the Favorable CV effects with these drugs pre- observed for widely recommended thera- drug exposure is too short, and/or the sumably resulted from effects beyond sim- pies to prevent CVD such as lipid lowering early events may be less amenable to ply lowering glucose in patients with with statins, antihypertensive therapy, the intervention. advanced diabetes, except perhaps in and aspirin (89). These three trials, to- The best way to demonstrate benefit the trial testing semaglutide, during which gether with similar but more controver- might be to select lower-risk patients who a pronounced A1C reduction occurred sial findings in the CANVAS Program, have not yet manifested CVD to find out in the group receiving active therapy. support an early and evolving case for whether their likelihood of developing Whether the beneficial effects of these the use of these antihyperglycemic ther- CVD can be reduced. To achieve a statisti- drugs would be similar or different if they apies in patients with type 2 diabetes and cally significant number of MACE events, were initiated closer to the time of diabetes evidence of clinical CVD who are .50 this will require larger and/or longer stud- diagnosis and earlier in the course of CVD years of age. ies but would yield valuable information development remains to be seen. Although the trials conducted thus far about CVD prevention. The current CVOT design also provides appear to both satisfy the FDA guidance only limited objective information about recommendations and provide additional Longer-Term Follow-Up Trial designs that prespecify longer-term harmful outcomes that occur .5 years useful information, there are clearly op- follow-up may be desirable to identify any after an intervention. Whether the new portunities for improvement (Table 4). longer-term safety issues and beneficial antidiabetes drugs currently being evalu- The next generation of diabetes trials effects that are either slowly evolving or ated have risks that may appear only after should be smarter, simpler, and innova- resulting from a legacy effect of earlier years of therapy is unknown, but CVOTs tively designed to make more efficient treatment. Toward that end, innovative with a duration of ,5 years are not likely use of resources and produce more gen- consent procedures that permit lifelong to detect them. Concerns may be greater eralizable results while still addressing follow-up should be explored, as should for therapeutic drugs with complex mech- safety issues. The design and conduct of strategies to increase patient adherence anisms of action that are not currently future trials should be standardized to and persistence with assigned therapy. well understood, including TZDs, DPP-4 in- enough to permit reliable between-trial The ability to capture adher- hibitors, and (despite their strong short- comparisons but also nimble enough to ence and relevant clinical and laboratory term protective effects) SGLT2 inhibitors. explore unanticipated safety concerns data will be crucial and should be facili- that may arise. In pursuing these goals, tated by use of more sophisticated elec- CVOTS OF THE FUTURE: some modifications to the current CVOT tronic medical records (EMRs). OPPORTUNITIES FOR noninferiority trial design, outlined be- Statistical approaches would need to IMPROVEMENT low, are worthy of consideration. be developed (and approved by regulatory As previously noted, the completed bodies) to permit robust assessment of CVOTs have yielded some important in- Lower-Risk and More Diverse key clinical outcomes over an extended formation about both safety concerns Populations timescale. For example, a trial could pro- and potential medical benefits. Regarding Incorporation of recent trial findings into vide definitive early information on the de- safety, the increased risk of HF hospitali- diabetes care guidelines has been limited gree of glucose lowering achieved (e.g., at zation with saxagliptin in SAVOR-TIMI to date, in part because the populations 1 year) and also be powered to assess the 53 (6) and a nonsignificant trend with studied were selected for high CV risk. impact of an intervention on a primary aloglipitin in EXAMINE (7,59) resulted in Although this practice is consistent with clinical outcome at a later time (e.g., at the addition of new safety warnings to FDA recommendations and essential to 3 years). Continued follow-up thereafter the labels of those medications (64). demonstrate safety, it remains unclear could examine recurrent or less frequent Given these safety signals for HF, which whether results seen in high-risk patients clinical outcomes, maximize collection of were unexpected from the registration are translatable to patients with a shorter safety data, and better assess the durabil- trials, it is likely that studies of individual duration of diabetes or without estab- ity of the glucose-lowering effect. antidiabetes drugs will continue to be lished CV complications. Even a few such trials could yield a rich required. Enrollment of very high-risk subjects harvest of clinically relevant information to Further, four of the completed tri- also may hamper the ability of trials to better inform clinical guidelines and ther- als (LEADER, SUSTAIN-6, EMPA-REG detect a benefit from drug therapy. As apeutic choices for patients and produce OUTCOME, and the CANVAS Program) examples, the ELIXA trial of lixisenatide more robust cost-effectiveness data for reported clinically relevant relative risk (9) and the EXAMINE trial of alogliptin health care providers. Longer trials would reductions of MACE outcomes in patients (7) exclusively recruited patients with a also provide a better understanding of the withhighCVrisk(Table 5) (10,11,13,14,60). recent hospitalization for an acute coro- evolution of treatment requirements over Although it is inherently inappropriate to nary event. Such patients are easy to time, especially if participants are random- compare results among trials because identify based on hospital records and ized to specific therapeutic sequences. they are conducted in different popula- are likely to have a high number of fu- tions, with different protocols, and at dif- ture MACE events. Thus, such trials may Active Comparators ferent sites, the numbers needed to treat be completed more quickly than those Another difficulty in interpreting and ap- (NNTs) to prevent a MACE occurrence in enrolling patients with more stable CVD. plying the results of most CVOTs com- EMPA-REG OUTCOME (63 over 3.1 years), However, this strategy may fail to detect a pleted to date concerns comparing the 26 Cardiovascular Outcomes Trials in Type 2 Diabetes Diabetes Care Volume 41, January 2018

Table 5—Risk reduction in four completed trials showing evidence of CV benefit LEADER (10) SUSTAIN-6 (11) EMPA-REG OUTCOME (13,60) CANVAS Program (14) Subjects (n) 9,340 3,297 7,020 10,142 Mean age (years) 64.3 64.6 63.1 63.3 Diabetes duration (years)* 12.8 13.9 57% .10 13.5 Mean baseline A1C (%) 8.7 8.7 8.1 8.2 Mean placebo-corrected A1C difference (%)† 20.4 20.7(0.5mgdose) 20.24 (10 mg dose) 20.58 21.0(1.0mgdose) 20.36 (25 mg dose) Median follow-up duration (years) 3.8 2.1 3.1 2.4 3-point MACE RRR (%) 13 26 14 14 3-point MACE ARR (%) 1.9 2.3 1.6 d‡ CV death RRR (%) 22 2 38 4§; 13| Nonfatal MI RRR (%) 12 26 13 15 Nonfatal stroke RRR (%) 11 39 +24 10 All-cause mortality RRR (%) 15 +5 32 13§; 10| HF hospitalization RRR (%) 13 +11 35 33 Worsening nephropathy RRR (%)¶ 22 36 39 40 Boldface type indicates statistical significance. +, increased relative risk; d, not reported; ARR, absolute risk reduction; RRR, relative risk reduction. *Reported as mean in all trials except EMPA-REG OUTCOME, which reported percentage of population with diabetes duration .10 years. †For LEADER and the CANVAS Program, difference was updated over time; for SUSTAIN-6 and EMPA-REG OUTCOME, difference was at study end. ‡Data needed to calculate ARR in 3-point MACE (specifically the NNT) were not reported for the CANVAS Program; instead this trial reported an incidence rate difference of 4.6 per 1,000 patient-years for canagliflozin vs. placebo. §Truncated integrated data set (refers to pooled data from CANVAS after 20 November 2012 plus CANVAS-R; prespecified in treating hierarchy as the principal data set for analysis for superiority of all-cause mortality and CV death in the CANVAS Program). |Nontruncated integrated data (refers to pooled data from CANVAS, including before 20 November 2012 plus CANVAS-R. ¶Worsening nephropathy is defined as the new onset of macroalbuminuria (urine albumin creatinine ratio .300 mg/g) or a doubling of the serum creatinine level and an eGFR rate of #45 mL/min/1.73 m2, the need for continuous renal-replacement therapy, or death from renal disease in LEADER, SUSTAIN, and EMPA-REG OUTCOME and as 40% reduction in eGFR, renal-replacement therapy, or death from renal causes in CANVAS. This outcome was a prespecified exploratory adjudicated outcome in LEADER, SUSTAIN-6, and CANVAS but not in EMPA-REG OUTCOME.

drug in question to placebo, with both tested to explore whether the CV benefits collection of additional clinical outcomes groups attempting to attain comparable are compounded. Between-trial consis- relevant to diabetes that are not collected glycemic control using regimens that in- tency in study design, enrollment criteria, routinely in current trials because of cost clude other medications. There are sev- and ascertainment of clinical data will fa- and complexity. Such outcomes include eral problems with this design. First, in cilitate these efforts. rates of cancer, bone fractures, depression, general, comparable glycemic control is Innovative Designs and dementia, as well as unexpected off- not achieved, with the study drug group Creative trial designs likely will be needed target side effects undetected during con- usually having lower A1C values. Such dif- to fully address current concerns. In addi- ventional drug development programs. This ferences might contribute to any CV ben- tion to traditional RCTs, factorial or adap- would extend assessment of safety beyond efits seen. Second, some drugs used more tive designs, superiority trials, trials CV safety and include long-term durability frequently in the comparison groups may embedded within health care systems or of effects. Sample sizes could be much themselves have CV effects. Third, mov- networks, and/or employment of “big larger using the embedded trial approach ing forward, it may not, for ethical rea- data” to dissect the effects of new diabe- because of lower costs per patient, enabling sons, be possible to restrict background tes medications may provide practical more powerful assessments of the hetero- use of drugs that have recently been opportunities for further investigation. geneity of treatment effects. Trials of this shown to have beneficial effects. Factorial designs would increase efficiency nature also could be crucial in realizing the However, using an active comparator by allowing simultaneous testing of multi- goal of precision pharmacotherapy in type 2 as a control will require sufficient knowl- ple interventions, and superiority trials diabetes, should different classes of drugs edge of the CV impact of the comparator would put greater emphasis on potential or specific drugs within classes be found to to avoid confounding interpretation of benefits of therapy and likely make re- be better for patients with different clinical the results. Although challenging, this cruitment easier by increasing the value characteristics (31). However, defining the may become feasible as our understand- proposition for participants. Pragmatic de- randomization process to avoid con- ing of the CV safety of newer antidiabetes signs that embed studies within health founders and any imbalances between drugs increases. Future trials may success- care systems’ EMR platforms would pro- the groups studied will be crucial. fully assess the impact of a new drug com- vide a natural conduit for conducting long- Although registries or EMRs may af- pared with a previously tested drug if the term posttrial observational follow-up ford the opportunity for large, pragmatic, newer trial enrolls a patient population studies. However, lessons from EXSCEL and relatively inexpensive trials, the suc- and adopts a design that is similar to (12) must be considered to avoid having cess of such trials likely will depend on that with which the comparator drug was low treatment persistence in such designs. having a simple intervention and an ap- studied. Combinations of drugs known to Embedding RCTs into routine clinical care propriate health system with an adequate be cardioprotective may also need to be would also enable the multidimensional EMR, as is the case with the ongoing care.diabetesjournals.org Cefalu and Associates 27

ADAPTABLE (Aspirin Dosing: A Patient- considered, especially when the statistical sponsor, is cumbersome and expensive. Centric Trial Assessing Benefits and Long- findings are robust, biologically plausible, Strategies to enhance the efficiency of term Effectiveness) trial of aspirin therapy and supportable by external evidence. trials, such as collecting CV outcomes for atherosclerotic CVD (90). As men- This way forward would require buy-in data from trials designed for other pur- tioned above, irrespective of trial design, from regulatory bodies, as well as mech- poses, should be strongly considered. Al- treatment randomization will remain cru- anisms to ensure equity for developers of though evidence that the expense of such cial to determining a medication’s effect first-in-class interventions and/or those studies is suppressing innovation in new without undue bias. who willingly adopt more complex trial diabetes drug development is lacking, it is designs. conceivable that, if the cost of CVOTs fi Standardized De nitions Another challenge is how best to incor- were reduced, more innovative research Safety and efficacy comparisons would be porate predefined safety concerns into could be supported. Thus, models for greatly enhanced by standardizing the the primary analysis. One possibility is cost-sharing among pharmaceutical, gov- definitions of important outcomes. The to add safety issues as components of ernmental, and other organizations strategies and definitions used to identify the primary composite outcome if the in- should be explored. events such as pancreatitis, pancreatic cluded safety and efficacy components malignancy, progression of retinopathy, can be weighted appropriately to reflect Involvement of Patients and Advocacy and worsening of renal function have var- their clinical importance. Such a com- Organizations ied considerably among the CVOTs con- bined end point would then reflect the When designing future trials, all stake- ducted to date. Collaborative efforts intervention’s overall net clinical benefit holders should be involved. The direct should be made to generate definitions to the patient. If there is adequate power, involvement of patients and their advo- for high-priority safety and microvascular it may be more desirable to evaluate key cates will be vitally important to ensure outcomes, akin to those established for efficacy and key safety outcomes sepa- that patients’ views and wishes are taken CV outcomes. rately as predefined coprimary end points, into account and that patient-related out- Modification of End Points and with superiority and noninferiority analy- come measures are fully integrated. Analyses ses respectively, as was done in EXSCEL There is some concern, however, that In light of the noted differences in effects (12). In general, however, the FDA typically patient-reported outcome measures are of drugs on components of MACE com- prefers trials to establish efficacy and not standardized, limiting comparisons posite end points, it may be desirable to safety separately, with approval based between studies. Efforts to improve reassess the end points and primary anal- on whether efficacy (benefit) outweighs patients’ understanding of clinical trials yses to be adopted in future studies to safety (harm). and to enlist the help of patients in de- provide a more nuanced look at CV out- veloping trial information literature and comes. Trials likely will still be powered Establishment of Biorepositories consent processes might improve pa- Many different mechanisms contribute to based on the time to the first primary end tients’ willingness to participate as sub- the occurrence of CV events, including point, despite disadvantages that include jects. The importance of minimizing atherosclerosis, hemodynamic changes, possible heterogeneity of the composite participant discontinuation, improving arrhythmias, and, for strokes, hemorrhagic components, results driven by components study drug adherence and treatment mechanisms. To gain additional insight of lesser importance, and insufficient persistence, and avoiding missing data into both mechanisms and subgroups, it power to draw definitive conclusions for cannot be overstated when seeking to would be desirable for future trials to all components of the composite (91). To obtain definitive trial outcomes. The rapid have informed consent to store biological address these limitations, alternative an- growth of digital health and monitoring (i.e., serum, urine, fecal, and DNA) samples alytic strategies such as weighted com- devices with Internet connectivity may of- in case unexpected results warrant further posite outcomes (92) and the win ratio fer further opportunities to access patient- investigation or for additional hypothesis- approach (93) should be considered. To specific outcomes or provide mechanistic generating studies. Greater use of patient better capture the full impact of treat- insights (29,96). phenotyping, genotyping, and metabolomic ments, subsequent and recurrent events data, similar to what is now being done in the also need to be evaluated for both bene- CONCLUSIONS field of oncology, may help to better identify fits and risks using validated statistical ap- The 2008 FDA guidance for diabetes drug predictors of benefitorharmthatsofar proaches (94,95). When testing drugs development greatly increased the collec- have escaped detection. Biorepositories within the same class, data from previous tion of data on conventional CV outcomes will be of key importance in this regard, trials should be used to inform both the of treatment. Studies completed to date as well, by allowing for subsequent bio- design and sample size requirements, and those now under way are providing marker, gut microbiota, and genomic anal- with a Bayesian approach considered to strong evidence for CV benefit for several yses to facilitate future precision medicine potentially limit sample sizes to more drugs and reassurance about the lack of opportunities. manageable and less costly proportions CV risk for many others. In addition, they without compromising study power. Im- Enhanced Efficiency and Cost-Sharing have yielded further information on portant secondary outcomes, especially Options non-CV benefits, such as reduction of re- mortality or serious irreversible morbidity The current system for evaluating the CV nal disease, and potential harms, such as events, should not be dismissed even if safety of new diabetes drugs, in which HF and lower-limb amputations. This the primary composite outcomes are not each is assessed individually in a large, information has built a strong founda- achieved; they should be independently long-term CVOT funded by a pharmaceutical tion of outcomes evidence that is truly 28 Cardiovascular Outcomes Trials in Type 2 Diabetes Diabetes Care Volume 41, January 2018

transforming diabetes care. The informa- grant support from AstraZeneca, Boehringer Ingel- and has received research support through his tion on improved CV outcomes for specific heim, and Novo Nordisk. J.S.S. has been an advisor institution from AstraZeneca, Eli Lilly, and Novo antidiabetes drugs should be considered to Adocia, AstraZeneca, BD Technologies, Boehringer Nordisk; these potential dualities have been re- Ingelheim, Dance Biopharm, Diavacs, Elcelyx Ther- viewed and managed by Oregon Health & Science in revised clinical treatment recommenda- apeutics, Eli Lilly, Ideal Life, ImmunoMolecular University. No other potential conflicts of interest tions, given that cardioprotection is an Therapeutics, Intarcia Therapeutics, Intrexon, relevant to this article were reported. added benefit. Merck, Orgenesis, Sanofi, Servier, vTv Therapeu- However, after nearly a decade of ex- tics, Valeritas, and Viacyte. He has received re- References search funding from Mesoblast and Viacyte. He perience with CVOTs designed on the ba- 1. U.S. Food and Drug Administration. Guidance is a member of the board of directors of Dexcom, for industry: diabetes mellitusdevaluating sis of the 2008 guidance, it is time to Intarcia Therapeutics, Moerae Matrix, and Vaso- cardiovascular risk in new antidiabetic thera- reevaluate how best to conduct such tri- Prep Surgical. He is a stock shareholder in Dance pies to treat type 2 diabetes [Internet]. Avail- fi Biopharm, Dexcom, Ideal Life, Intarcia Therapeu- als. Future trials might bene tfromfocus- able from www.fda.gov/downloads/Drugs/ tics, Intrexon, Moerae Matrix, and VasoPrep Sur- ing on populations more typical of those GuidanceComplianceRegulatoryInformation/ gical. J.B.G. is a consultant to Boehringer Ingelheim seen in routine care, with longer duration Guidances/ucm071627.pdf. Accessed 31 October and Daiichi Sankyo and has received research fund- 2016 of follow-up, greater consistency in the ing from AstraZeneca, GlaxoSmithKline, and Merck. ascertainment of outcomes, and improved 2. Laakso M. Cardiovascular disease in type 2 di- J.B.B. has received contracted consulting fees, paid abetes from population to man to mechanisms: statistical methods for analysis. In addi- to his institution, and travel support from Adocia, the Kelly West Award Lecture 2008. Diabetes Care tion, greater efficiency and relevance to AstraZeneca, Dance Biopharm, Dexcom, Elcelyx 2010;33:442–449 clinical practice might be obtained by Therapeutics, Eli Lilly, Fractyl, GI Dynamics, Intar- 3. Nissen SE, Wolski K. Effect of rosiglitazone on cia Therapeutics, Lexicon, Merck, Metavention, collaboration of various stakeholders in the risk of myocardial infarction and death from NovaTarg, Novo Nordisk, Orexigen, PhaseBio, cardiovascular causes. N Engl J Med 2007;356: d fi health care including pharmaceutical Sano , Shenzhen HighTide, and vTv Thera- 2457–2471 companies, regulatory groups, insurers, peutics and grant support from AstraZeneca, 4. Caveney E, Turner JR. White paper. A review of health systems, consumer advocates, Bayer, Boehringer Ingelheim, Eli Lilly, GI Dy- FDA guidance: understanding the FDA guidance namics, GlaxoSmithKline, Intarcia Therapeutics, and people with diabetes themselvesdin on assessing cardiovascular risks for new antidia- Johnson & Johnson, Lexicon, Medtronic, Merck, betic therapies [Internet]. Available from http:// the conduct of trials, perhaps by embed- fi Novo Nordisk, Orexigen, Sano ,ScionNeuroStim, www.quintiles.com/library/white-papers/new- ding clinical protocols within health sys- Takeda, and Theracos. He has received fees and fda-guidance-on-antidiabetic-therapies. Accessed tems. It is to be expected that future holds stock options in Insulin Algorithms and 31 October 2016 trials, like those already completed, will PhaseBioandservesontheboardofthe 5. European Medicines Agency. Guideline on clin- AstraZeneca HealthCare Foundation. S.E.I. has served provide both answers to the primary re- ical investigation of medicinal products in the as a consultant to Alere, Janssen, and vTv Thera- treatment or prevention of diabetes mellitus [In- search questions posed and new insights peutics. He has participated on clinical trial steer- fi ternet], 2012. Available from http://www.ema regarding previously unexpected bene ts ing or executive committees for AstraZeneca, .europa.eu/docs/en_GB/document_library/ and risks. Boehringer Ingelheim, Daiichi Sankyo, Eisai, Scientific_guideline/2012/06/WC500129256.pdf. Novo Nordisk, and Sanofi/Lexicon. He has served Accessed 3 November 2016 as a member of a data monitoring committee for 6. Scirica BM, Bhatt DL, Braunwald E, et al.; Intarcia Therapeutics. L.A.L. has received research SAVOR-TIMI 53 Steering Committee and Investi- Acknowledgments. Writing and editing sup- funding from, has provided continuing medica- gators. Saxagliptin and cardiovascular outcomes port services for this article were provided by tion education on behalf of, and/or has acted as in patients with type 2 diabetes mellitus. N Engl J Debbie Kendall of Kendall Editorial in Richmond, an advisor to AstraZeneca, Boehringer Ingelheim, Med 2013;369:1317–1326 Virginia. The authors thank Christian S. Kohler, Eli Lilly, GlaxoSmithKline, Janssen, Merck, Novo 7. White WB, Cannon CP, Heller SR, et al.; the ADA’s Associate Publisher, Scholarly Journals, Nordisk, Sanofi, and Servier. I.R. has received fees EXAMINE Investigators. Alogliptin after acute cor- and his staff for their assistance, guidance, and for serving on advisory boards for AstraZeneca, onary syndrome in patients with type 2 diabetes. expertise in convening the Expert Forum. Boehringer Ingelheim, Concenter BioPharma/ N Engl J Med 2013;369:1327–1335 Duality of Interest. W.T.C., prior to joining the Silkim, Eli Lilly, LabStyle Innovations, Merck Sharp 8. Green JB, Bethel MA, Armstrong PW, et al.; ADA, received grant support awarded to his & Dohme, Novo Nordisk, Orgenesis, Pfizer, Sanofi, TECOS Study Group. Effect of sitagliptin on cardio- institution from Novo Nordisk and Sanofi.He and SmartZyme Innovation; fees for consultant vascular outcomes in type 2 diabetes. N Engl J served as a consultant for Adocia, Intarcia Ther- services from AstraZeneca/Bristol-Myers Squibb, Med 2015;373:232–242 apeutics, and Sanofi. He has no active disclo- Diabetes Medical Center, FutuRx, Gili Medical, 9. Pfeffer MA, Claggett B, Diaz R, et al.; ELIXA sures. S.K. is a consultant to and advisor for Insuline Medical, Kamada, and Medial EarlySign; Investigators. Lixisenatide in patients with type 2 Boehringer Ingelheim, Eli Lilly, and Novo Nordisk and fees for serving on speakers’ bureaus from diabetes and acute coronary syndrome. N Engl J and is a stock shareholder in Johnson & Johnson. AstraZeneca/Bristol-Myers Squibb, Boehringer In- Med 2015;373:2247–2257 H.C.G. holds the McMaster-Sanofi Population gelheim, Eli Lilly, Johnson & Johnson, Merck Sharp 10. Marso SP, Daniels GH, Brown-Frandsen K, Health Institute Chair in Diabetes Research and & Dohme, Novartis Pharma AG, Novo Nordisk, et al.; LEADER Steering Committee; LEADER Trial Care. He has received research grant support Sanofi, and Teva. He is a stock shareholder in Investigators. Liraglutide and cardiovascular out- from AstraZeneca, Eli Lilly, Merck, and Sanofi; GlucoMe, Insuline Medical, LabStyle Innovations, comes in type 2 diabetes. N Engl J Med 2016;375: honoraria for speaking engagements from and Orgenesis. J.R. has served on scientificadvi- 311–322 AstraZeneca, Boehringer Ingelheim, Novo Nordisk, sory boards and received honoraria or consulting 11. Marso SP, Bain SC, Consoli A, et al.; SUSTAIN- and Sanofi; and consulting fees from Abbott, Amgen, fees from AstraZeneca, Boehringer Ingelheim, Eli 6 Investigators. Semaglutide and cardiovascular AstraZeneca, Boehringer Ingelheim, Eli Lilly, Merck, Lilly, Intarcia Therapeutics, Janssen, Novo Nordisk, outcomes in patients with type 2 diabetes. N Engl Novo Nordisk, and Sanofi. R.R.H. reports receiving and Sanofi. He has received grants/research sup- J Med 2016;375:1834–1844 grants from AstraZeneca; grants and personal fees port from AstraZeneca, Boehringer Ingelheim, 12. Holman RR, Bethel MA, Mentz RJ, et al.; from Bayer, Boehringer Ingelheim, and Merck; Bristol-Myers Squibb, Eli Lilly, Genentech, EXSCEL Study Group. Effects of once-weekly exe- personal fees from Amgen, Novartis, and Servier; GlaxoSmithKline, Intarcia Therapeutics, Janssen, natide on cardiovascular outcomes in type 2 di- and other support from Elcelyx Therapeutics, Lexicon, Merck, Novo Nordisk, Pfizer, and Sanofi. abetes. N Engl J Med 2017;377:1228–1239 GlaxoSmithKline, Janssen, and Takeda.B.Z. isa con- M.C.R. has received honoraria for consulting, 13. Zinman B, Wanner C, Lachin JM, et al.; EMPA- sultant to and has received honoraria from speaking, or serving on committees for Adocia, REG OUTCOME Investigators. Empagliflozin, car- AstraZeneca, Boehringer Ingelheim, Eli Lilly, Janssen, AstraZeneca, Elcelyx Therapeutics, Eli Lilly, diovascular outcomes, and mortality in type 2 Merck, Novo Nordisk, and Sanofi and has received GlaxoSmithKline, Sanofi, Theracos, and Valeritas diabetes. N Engl J Med 2015;373:2117–2128 care.diabetesjournals.org Cefalu and Associates 29

14. Neal B, Perkovic V, Mahaffey KW, et al.; treatment of type 1 diabetes and long-term mor- 204042Orig1s000SumR.pdf. Accessed 5 May CANVAS Program Collaborative Group. Canagliflozin tality. JAMA 2015;313:45–53 2017 and cardiovascular and renal events in type 2 di- 28. Holman RR, Paul SK, Bethel MA, Matthews 42. Regier EE, Venkat MV, Close KL. More than abetes. N Engl J Med 2017;377:644–657 DR, Neil HAW. 10-year follow-up of intensive glu- seven years of hindsight: revisiting the FDA’s 15. Rawshani A, Rawshani A, Franzen´ S, et al. cose control in type 2 diabetes. N Engl J Med 2008; 2008 guidance on cardiovascular outcomes trials Mortality and cardiovascular disease in type 1 359:1577–1589 for type 2 diabetes medications. Clin Diabetes and type 2 diabetes. N Engl J Med 2017;376: 29. Hayward RA, Reaven PD, Wiitala WL, et al.; 2016;34:173–180 1407–1418 VADT Investigators. Follow-up of glycemic control 43. Novo Nordisk. Novo Nordisk receives com- 16. National Cholesterol Education Program and cardiovascular outcomes in type 2 diabetes. plete response letter in the US for Tresiba and (NCEP) Expert Panel on Detection, Evaluation, N Engl J Med 2015;372:2197–2206 Ryzodeg [Internet]. Available from https://www and Treatment of High Blood Cholesterol in Adults 30. ACCORD Study Group. Nine-year effects of .novonordisk.com/bin/getPDF.1676900.pdf. (Adult Treatment Panel III). Third Report of the 3.7 years of intensive glycemic control on cardio- Accessed 4 November 2016 National Cholesterol Education Program (NCEP) vascular outcomes. Diabetes Care 2016;39:701– 44. Boehringer Ingelheim. CAROLINA: cardio- Expert Panel on Detection, Evaluation, and Treat- 708 vascular outcome study of linagliptin versus ment of High Blood Cholesterol in Adults (Adult 31. Holman RR, Sourij H, Califf RM. Cardiovascu- glimepiride in patients with type 2 diabetes. In: Treatment Panel III) final report. Circulation 2002; lar outcome trials of glucose-lowering drugs or ClinicalTrials.gov [Internet]. Bethesda, MD, Na- 106:3143–3421 strategies in type 2 diabetes. Lancet 2014;383: tional Library of Medicine. Available from https:// 17.NathanDM,GenuthS,LachinJ,etal.;Diabe- 2008–2017 clinicaltrials.gov/show/NCT01243424. NLM Identifier: tes Control and Complications Trial Research 32. John M, Gopalakrishnan Unnikrishnan A, NCT01243424. Accessed 6 April 2017 Group. The effect of intensive treatment of dia- Kalra S, Nair T. Cardiovascular outcome trials 45. Boehringer Ingelheim. Cardiovascular and re- betes on the development and progression of for anti-: a holy grail of nal microvascular outcome study with linagliptin long-term complications in insulin-dependent di- drug development? Indian Heart J 2016;68: in patients with type 2 diabetes mellitus abetes mellitus. N Engl J Med 1993;329:977–986 564–571 (CARMELINA). In: ClinicalTrials.gov [Internet]. Be- 18. UK Prospective Diabetes Study (UKPDS) 33. Nissen SE, Wolski K, Topol EJ. Effect of mur- thesda, MD, National Library of Medicine. Available Group. Intensive blood-glucose control with sul- aglitazar on death and major adverse cardiovas- from https://clinicaltrials.gov/show/NCT01897532. phonylureas or insulin compared with conven- culareventsinpatientswithtype2diabetes NLM Identifier: NCT01897532. Accessed 10 tional treatment and risk of complications in mellitus. JAMA 2005;294:2581–2586 February 2017 patients with type 2 diabetes (UKPDS 33). Lancet 34. Drugs.com. Bristol-Myers Squibb and Merck 46. Janssen Research & Development. Evaluation 1998;352:837–853 joint statement on FDA advisory committee vote of the effects of canagliflozin on renal and cardio- 19. UK Prospective Diabetes Study (UKPDS) on Pargluva (muraglitazar), an investigational vascular outcomes in participants with diabetic Group. Effect of intensive blood-glucose control oral treatment for type 2 diabetes [Internet]. nephropathy (CREDENCE). In: ClinicalTrials.gov with metformin on complications in overweight Available from https://www.drugs.com/nda/ [Internet]. Bethesda, MD, National Library of patients with type 2 diabetes (UKPDS 34). Lancet pargluva_050909.html. Accessed 3 November Medicine. Available from https://clinicaltrials 1998;352:854–865 2016 .gov/show/NCT02065791. NLM Identifier: 20. Gerstein HC, Miller ME, Byington RP, et al.; 35. Home PD, Pocock SJ, Beck-Nielsen H, et al.; NCT02065791. Accessed 10 February 2017 Action to Control Cardiovascular Risk in Diabetes RECORD Study Team. Rosiglitazone evaluated for 47. AstraZeneca. Multicenter trial to evaluate the Study Group. Effects of intensive glucose lowering in cardiovascular outcomes in oral agent combina- effect of dapagliflozin on the incidence of cardio- type 2 diabetes. N Engl J Med 2008;358:2545–2559 tion therapy for type 2 diabetes (RECORD): a mul- vascular events (DECLARE-TIMI 58). In: Clinical- 21. Patel A, MacMahon S, Chalmers J, et al.; ticentre, randomised, open-label trial. Lancet Trials.gov [Internet]. Bethesda, MD, National ADVANCE Collaborative Group. Intensive blood 2009;373:2125–2135 Library of Medicine. Available from https:// glucose control and vascular outcomes in patients 36. Mahaffey KW, Hafley G, Dickerson S, et al. clinicaltrials.gov/show/NCT01730534. NLM Iden- with type 2 diabetes. N Engl J Med 2008;358: Results of a reevaluation of cardiovascular out- tifier: NCT01730534. Accessed 10 February 2017 2560–2572 comes in the RECORD trial. Am Heart J 2013; 48. Merck Sharp & Dohme. Cardiovascular out- 22. Duckworth W, Abraira C, Moritz T, et al.; 166:240–249.e1 comes following ertugliflozin treatment in type 2 di- VADT Investigators. Glucose control and vascular 37. U.S. Food and Drug Administration. FDA drug abetes mellitus participants with vascular disease, the complications in veterans with type 2 diabetes. safety communication: FDA requires removal of VERTIS CV study (MK-8835-004). In: ClinicalTrials.gov N Engl J Med 2009;360:129–139 some prescribing and dispensing restrictions for [Internet]. Bethesda, MD, National Library of Medi- 23. Ohkubo Y, Kishikawa H, Araki E, et al. Inten- rosiglitazone-containing diabetes medicines [In- cine.Availablefromhttps://clinicaltrials.gov/show/ sive insulin therapy prevents the progression of ternet]. Available from https://www.fda.gov/ NCT01986881. NLM Identifier: NCT01986881. diabetic microvascular complications in Japanese drugs/drugsafety/ucm376389.htm. Accessed 9 Accessed 14 February 2017 patients with non-insulin-dependent diabetes November 2017 49. AstraZeneca. Study to evaluate the effect of mellitus: a randomized prospective 6-year study. 38. Dormandy JA, Charbonnel B, Eckland DJ, dapagliflozin on the incidence of worsening heart Diabetes Res Clin Pract 1995;28:103–117 et al.; PROactive Investigators. Secondary preven- failure or cardiovascular death in patients with 24. Ray KK, Seshasai SR, Wijesuriya S, et al. Effect tion of macrovascular events in patients with chronic heart failure (Dapa-HF). In: ClinicalTrials of intensive control of glucose on cardiovascular type 2 diabetes in the PROactive Study (PROspec- .gov [Internet]. Bethesda, MD, National Library outcomes and death in patients with diabetes tive pioglitAzone Clinical Trial In macroVascular of Medicine. Available from https://clinicaltrials mellitus: a meta-analysis of randomised con- Events): a randomised controlled trial. Lancet .gov/show/NCT03036124. NLM Identifier: trolled trials. Lancet 2009;373:1765–1772 2005;366:1279–1289 NCT03036124. Accessed 6 April 2017 25. Turnbull FM, Abraira C, Anderson RJ, et al.; 39. Kernan WN, Viscoli CM, Furie KL, et al.; IRIS 50. AstraZeneca. A study to evaluate the effect of Control Group. Intensive glucose control and mac- Trial Investigators. Pioglitazone after ischemic dapagliflozin on renal outcomes and cardiovascu- rovascular outcomes in type 2 diabetes. Diabeto- stroke or transient ischemic attack. N Engl J Med lar mortality in patients with chronic kidney dis- logia 2009;52:2288–2298 2016;374:1321–1331 ease (Dapa-CKD). In: ClinicalTrials.gov [Internet]. 26. Nathan DM, Cleary PA, Backlund JY, et al.; 40. Young LH, Viscoli CM, Curtis JP, et al. IRIS Bethesda, MD, National Library of Medicine. Diabetes Control and Complications Trial/ Investigators. Cardiac outcomes after ischemic Available from https://clinicaltrials.gov/show/ Epidemiology of Diabetes Interventions and Com- stroke or TIA: effects of pioglitazone in patients NCT03036150. NLM Identifier: NCT03036150. plications (DCCT/EDIC) Study Research Group. In- with insulin resistance without diabetes. Circula- Accessed 4 April 2017 tensive diabetes treatment and cardiovascular tion 2017;135:1882–1893 51. Boehringer Ingelheim. Empagliflozin out- disease in patients with type 1 diabetes. N Engl J 41. U.S. Food and Drug Administration Center come trial in patients with chronic heart failure Med 2005;353:2643–2653 for Drug Evaluation and Research. Application with reduced ejection fraction (EMPEROR- 27. Orchard TJ, Nathan DM, Zinman B, et al.; Number 204042Orig1s000: summary re- Reduced). In: ClinicalTrials.gov [Internet]. Bethesda, Writing Group for the DCCT/EDIC Research view [Internet]. Available from https://www MD, National Library of Medicine. Available from Group. Association between 7 years of intensive .accessdata.fda.gov/drugsatfda_docs/nda/2013/ https://clinicaltrials.gov/show/NCT03057977. 30 Cardiovascular Outcomes Trials in Type 2 Diabetes Diabetes Care Volume 41, January 2018

NLM Identifier: NCT03057977. Accessed 6 April omarigliptin in patients with type 2 diabetes mel- the analysis strategy for the CANVAS Program: a 2017 litus. Cardiovasc Diabetol 2017;16:112 prespecified plan for the integrated analyses of 52. Boehringer Ingelheim. Empagliflozin out- 63. Merck. Merck provides update on filing plans the CANVAS and CANVAS-R trials. Diabetes Obes come trial in patients with chronic heart failure for omarigliptin, an investigational DPP-4 inhibitor Metab 2017;19:926–935 withpreservedejectionfraction(EMPEROR- for type 2 diabetes [Internet]. Available from 76. Victoza liraglutide injection 1.2 mg/1.8 mg Preserved). In: ClinicalTrials.gov [Internet]. http://www.mrknewsroom.com/news/company- [prescribing information]. Plainsboro, NJ, Novo Bethesda, MD, National Library of Medicine. statements/merck-provides-update-filing-plans- Nordisk, 2017 Available from https://clinicaltrials.gov/show/ omarigliptin-investigational-dpp-4-inhibi. 77. Novo Nordisk. Semaglutide receives positive NCT03057951. NLM Identifier: NCT03057951. Accessed 7 April 2017 16-0 vote in favour of approval from FDA Advisory Accessed 6 April 2017 64. U.S. Food and Drug Administration. FDA drug Committee [Internet]. Available from https:// 53. Intarcia Therapeutics. Intarcia announces safety communication: FDA adds warnings about www.novonordisk.com/bin/getPDF.2142854.pdf. successful cardiovascular safety results in phase heart failure risk to labels of type 2 diabetes med- Accessed 19 October 2017 3 FREEDOM-CVO trial for ITCA 650, an investiga- icines containing saxagliptin and alogliptin [Inter- 78. Intarcia Therapeutics. Intarcia submits new tional therapy for type 2 diabetes: company also net]. Available from http://www.fda.gov/Drugs/ drug application (NDA) to FDA for U.S. marketing reports new $75 million financing for manufactur- DrugSafety/ucm486096.htm. Accessed 1 January approval of ITCA 650 in type 2 diabetes [Internet]. ing scale-up and inventory build for anticipated 2017 Available from https://www.intarcia.com/media/ global launch of ITCA 650 [Internet]. Available 65. Rosenstock J, Marx N, Kahn SE, et al. Cardio- press-releases/2016-nov-21-nda-submission.html. from https://www.intarcia.com/media/press- vascular outcome trials in type 2 diabetes and the Accessed 10 April 2017 releases/2016-may-6-cardiovascular-safety.html. sulphonylurea controversy: rationale for the 79. Lincoff AM, Tardif J-C, Schwartz GG, et al.; Accessed 14 February 2017 active-comparator CAROLINA trial. Diab Vasc Dis AleCardio Investigators. Effect of aleglitazar on 54. Gerstein HC, Colhoun HM, Dagenais GR, Res 2013;10:289–301 cardiovascular outcomes after acute coronary et al.; REWIND Trial Investigators. Design and 66. Neal B, Perkovic V, Matthews DR, et al.; syndrome in patients with type 2 diabetes melli- baseline characteristics of participants in the Re- CANVAS-R Trial Collaborative Group. Rationale, de- tus: the AleCardio randomized clinical trial. JAMA searching cardiovascular Events with a Weekly sign and baseline characteristics of the CANagliflozin 2014;311:1515–1525 INcretin in Diabetes (REWIND) trial on the cardio- cardioVascular Assessment Study-Renal (CANVAS- 80. Erdmann E, Califf R, Gerstein HC, et al. Effects vascular effects of . Diabetes Obes R): a randomized, placebo-controlled trial. Diabetes of the dual peroxisome proliferator-activated re- Metab 14 July 2017 [Epub ahead of print]. Obes Metab 2017;19:387–393 ceptor activator aleglitazar in patients with type 2 https://doi.org/10.1111/dom.13028 67. Eli Lilly. Jardiance (empagliflozin) tablets to be diabetes mellitus or prediabetes. Am Heart J 55. GlaxoSmithKline. Effect of albiglutide, studied in chronic kidney disease [Internet]. Avail- 2015;170:117–122 when added to standard blood glucose lower- able from https://investor.lilly.com/releasedetail. 81. Vaccaro O, Masulli M, Nicolucci A, et al.; Thia- ing therapies, on major cardiovascular events cfm?releaseid=1029854. Accessed 11 October zolidinediones Or Sulfonylureas Cardiovascular in subjects with type 2 diabetes mellitus. In: 2017 Accidents Intervention Trial (TOSCA.IT) study ClinicalTrials.gov [Internet]. Available from https:// 68. Ferrannini E, Mark M, Mayoux E. CV protec- group; Italian Diabetes Society. Effects on the clinicaltrials.gov/show/NCT02465515. NLM tion in the EMPA-REG OUTCOME trial: a “thrifty incidence of cardiovascular events of the addition Identifier: NCT02465515. Accessed 14 February substrate” hypothesis. Diabetes Care 2016;39: of pioglitazone versus sulfonylureas in patients 2017 1108–1114 with type 2 diabetes inadequately controlled 56. Novo Nordisk A/S. A trial investigating the 69. Mudaliar S, Alloju S, Henry RR. Can shift in with metformin (TOSCA.IT): a randomised, multi- cardiovascular safety of oral semaglutide in subjects fuel energetics explain the beneficial cardiorenal centre trial. Lancet Diabetes Endocrinol 2017;5: with type 2 diabetes (PIONEER 6). In: ClinicalTrials outcomes in the EMPA-REG OUTCOME study? A 887–897 .gov [Internet]. Available from https://clinicaltrials unifying hypothesis. Diabetes Care 2016;39: 82. Gerstein HC, Bosch J, Dagenais GR, et al.; .gov/show/NCT02692716. NLM Identifier: 1115–1122 ORIGIN Trial Investigators. Basal insulin and car- NCT02692716. Accessed 29 June 2017 70. Abdul-Ghani M, DeFronzo RA, Del Prato S, diovascular and other outcomes in dysglycemia. 57. Marso SP, McGuire DK, Zinman B, et al.; Chilton R, Singh R, Ryder REJ. Cardiovascular dis- N Engl J Med 2012;367:319–328 DEVOTE Study Group. Efficacy and safety of de- ease and type 2 diabetes: has the dawn of a new 83. ORIGIN Trial Investigators. Cardiovascular gludec versus glargine in type 2 diabetes. N Engl J era arrived? Diabetes Care 2017;40:813–820 and other outcomes postintervention with insulin Med 2017;377:723–732 71. U.S. Food and Drug Administration. FDA ap- glargine and omega-3 fatty acids (ORIGINALE). Di- 58. Holman RR, Coleman RL, Chan JCN, et al.; ACE proves Jardiance to reduce cardiovascular death abetes Care 2016;39:709–716 Study Group. Effects of acarbose on cardiovascu- in adults with type 2 diabetes [Internet]. Available 84. Canadian Diabetes Association Clinical lar and diabetes outcomes in patients with coro- from http://www.fda.gov/NewsEvents/Newsroom/ Practice Guidelines Expert Committee. Phar- nary heart disease and impaired glucose tolerance PressAnnouncements/ucm531517.htm. Accessed macologic management of type 2 diabetes: (ACE): a randomised, double-blind, placebo- 14 February 2017 2016 interim update. Can J Diabetes 2016;40: controlled trial. Lancet Diabetes Endocrinol 72. American Diabetes Association. Pharmaco- 484–486 2017;5:877–886 logic approaches to glycemic control. Sec. 8. In 85. McMurray JJ, Gerstein HC, Holman RR, 59. Zannad F, Cannon CP, Cushman WC, et al.; Standards of Medical Care in Diabetesd2017. Pfeffer MA. Heart failure: a cardiovascular out- EXAMINE Investigators. Heart failure and mortal- Diabetes Care 2017;40(Suppl. 1):S64–S74 come in diabetes that can no longer be ignored. ity outcomes in patients with type 2 diabetes tak- 73. Neal B, Perkovic V, de Zeeuw D, et al. Ratio- Lancet Diabetes Endocrinol 2014;2:843–851 ing alogliptin versus placebo in EXAMINE: a nale, design, and baseline characteristics of the 86. Fitchett D, Zinman B, Wanner C, et al.; EMPA- multicentre, randomised, double-blind trial. Lan- Canagliflozin Cardiovascular Assessment Study REG OUTCOMEÒ trial investigators. Heart failure cet 2015;385:2067–2076 (CANVAS)–a randomized placebo-controlled trial. outcomes with empagliflozin in patients with 60. Wanner C, Inzucchi SE, Lachin JM, et al.; Am Heart J 2013;166:217–223.e11 type 2 diabetes at high cardiovascular risk: results EMPA-REG OUTCOME Investigators. Empagliflozin 74. Janssen Pharmaceuticals. FDA briefing docu- of the EMPA-REG OUTCOMEÒ trial. Eur Heart J and progression of kidney disease in type 2 diabetes. ment: NDA 204042 Invokana (canagliflozin) tab- 2016;37:1526–1534 N Engl J Med 2016;375:323–334 lets [Internet]. Endocrinologic and Metabolic 87. Weidmann P, Ferrari P. Central role of sodium 61. Theodorakis MJ, Coleman RL, Feng H, et al. Drugs Advisory Committee Meeting. 10 January in hypertension in diabetic subjects. Diabetes Baseline characteristics and temporal differences 2013. Available from https://wayback.archive-it Care 1991;14:220–232 in Acarbose Cardiovascular Evaluation (ACE) trial .org/7993/20170405220010/https://www.fda.gov/ 88. Gerstein HC, Miller ME, Genuth S, et al.; participants. Am Heart J. 1 September 2017 [Epub downloads/AdvisoryCommittees/Committees ACCORD Study Group. Long-term effects of inten- ahead of print]. https://doi.org/10/1016/j.ahj MeetingMaterials/Drugs/Endocrinologicand sive glucose lowering on cardiovascular outcomes. .2017.09.001 MetabolicDrugsAdvisoryCommittee/UCM334550 N Engl J Med 2011;364:818–828 62. Gantz I, Chen M, Suryawanshi S, et al. A ran- .pdf. Accessed 10 October 2017 89. The NNT Group. Quick summaries of evidence- domized, placebo-controlled study of the cardio- 75. Neal B, Perkovic V, Mahaffey KW, et al.; based medicine [Internet]. Available from http:// vascular safety of the once-weekly DPP-4 inhibitor CANVAS Program collaborative group. Optimizing www.thennt.com. Accessed 5 February 2017 care.diabetesjournals.org Cefalu and Associates 31

90. Jones WS, Roe MT, Antman EM, et al. The 93. Pocock SJ, Clayton TC, Stone GW. Design of review of statistical methodology, with application changing landscape of randomized clinical trials major randomized trials: part 3 of a 4-part series to CHARM-Preserved. Eur J Heart Fail 2014;16: in cardiovascular disease. J Am Coll Cardiol on statistics for clinical trials. J Am Coll Cardiol 33–40 2016;68:1898–1907 2015;66:2757–2766 96. Gallin EK, Bond E, Califf RM, et al. Forging 91. Bethel MA, Holman R, Haffner SM, et al. De- 94. Rogers JK, Jhund PS, Perez AC, et al. Effect of stronger partnerships between academic health termining the most appropriate components for a rosuvastatin on repeat heart failure hospitaliza- centers and patient-driven organizations. Acad composite clinical trial outcome. Am Heart J 2008; tions: the CORONA Trial (Controlled Rosuvastatin Med 2013;88:1220–1224 156:633–640 Multinational Trial in Heart Failure). JACC Heart 97. Bergenstal RM, Bailey CJ, Kendall DM. Type 2 92. Stolker JM, Spertus JA, Cohen DJ, et al. Rethink- Fail 2014;2:289–297 diabetes: assessing the relative risks and benefits ing composite end points in clinical trials: insights from 95. Rogers JK, Pocock SJ, McMurray JJ, et al. Ana- of glucose-lowering medications. Am J Med 2010; patientsandtrialists. Circulation2014;130:1254–1261 lysing recurrent hospitalizations in heart failure: a 123:374.e9–e18