Cardiovascular Outcomes With Antihyperglycemic Therapy: Past, Present, and Future Impact on Practice

Faculty

Agenda

CV Outcomes in Type 2 Diabetes: Why Do We Care?

Complications of Diabetes^[1-8]

• Chronic complications lead to disability and death
• Traditional microvascular complications are affected by glycemic control and illness duration
• Death of patient with diabetes is commonly attributable to cardiovascular (CV) complications -- ie, stroke, cardiovascular disease (CVD)
• Heart failure (HF) outcomes might be impacted by treatment

Diabetes Increases the Risk of a CV Event^[9]

• Patients with diabetes have a 2-fold higher rate of traditional coronary heart disease complications -- ie, death, heart attack, any form of stroke

Patients With Diabetes Presenting With ACS Have Poor Prognosis^[10]

• Patients with diabetes have an approximately 50% increase in rate of death from any cause after having an acute coronary syndrome (ACS)
• The rate of cardiovascular death is increased by over 50%
• The event rate for congestive heart failure (CHF) is higher than that for any other complications after ACS in patients with diabetes

CVD Drives Healthcare Costs in Patients With Diabetes^[11]

• Large database in a managed care organization analyzed to examine healthcare costs associated with the management of comorbidities in patients with diabetes
• Patients with diabetes have dramatically higher healthcare costs associated with complications
• The cost of pharmaceutical agents pales in comparison to the cost of medical complications

UKPDS: HbA_1c and Rates for MI and Microvascular Complications^[12]

• In the UKPDS, for every 1% reduction in glycated hemoglobin (HbA_1c), there was an approximate 14% reduction in risk of CV endpoints
• In long-term CV trials, no impact of tight glycemic control on CV outcomes has been observed, especially in the randomized, controlled component of studies

Impact of Intensive Therapy in Diabetes in Major Clinical Trials^[13]

• The impact of intensive therapy on cardiovascular outcomes has been examined in a number of open-label observational studies
• Patients in these trials had 10 to 20 years of follow-up
• Reductions in rates of myocardial infarction (MI) were seen in the DCCT and UKPDS trials
• Reductions in rates of MI were not seen in the ACCORD or ADVANCE trials
• In recent long-term follow-up of the VADT study, intensive control of HbA_1c led to significant reduction in MI rates

VADT: Intensive Therapy Group: Relationship of Diabetes Duration and HR for CVD Events^[14]

• In the VADT study, a clear association between risk of MI and tight control of HbA_1c was observed
• Patients with diabetes duration <15 years experienced a reduction in risk of CVD

RACED Substudy of VADT: Effect of Glycemic Control on CVD Events by Baseline CAC Score^[15]

• The RACED substudy of VADT reflects results of the end of the randomized trial
• The substudy used coronary artery calcium (CAC) measured by computed tomography (CT), which correlated with atherogenic burden in coronary arteries
• Patients with CAC <100 had substantially fewer events than patients with scores >100

Atherosclerosis Timeline^[16]

• Endothelial dysfunction has the greatest impact on the development of atherosclerosis before atherogenesis occurs
• The UKPDS study enrolled patients before they had developed atherosclerosis
• The ACCORD, VADT, and ADVANCE studies started when patients already had significant fibrous plaque; glycemic control will not have much impact at that point
• Early intervention can make a different over a decade or more of life
• In later intervention, reducing risk by reducing HbA_1c is less important than considering other atherogenic risk factors

Mechanisms by Which Diabetes Leads to CHD^[17]

• Hypertension can lead to coronary heart disease (CHD)
• Infection can lead to increased inflammation
• The impact of hyperglycemia on advanced glycosylated end products and oxidative stress leads to atherosclerosis
• Increased inflammation, insulin resistance, hypertension, endothelial dysfunction, dyslipidemia, and prothrombotic state are related to visceral adiposity
• Atherosclerosis is subclinical until symptoms/and or an event occur

Overview of CV Outcome Studies With Antidiabetic Agents^[18]

Evolution of the FDA CV Safety Requirements and CV Safety Concerns^[19-22]

• Development of muraglitazar, a dual-α/γ agonist, was discontinued because of an increased risk of death, major CV events, CHF, and renal insufficiency in studies
• In 2007, a meta-analysis of studies of rosiglitazone suggested an increased CV risk
  • Withdrawn from major markets
  • Prescription restrictions in United States

FDA Postmarketing CVOT Requirements^[23]

• Companies marketing antidiabetic agents must provide data showing no increased CV risk
• Postmarketing CV outcome trials with a placebo comparator are required for agents with a hazard ratio (HR) between 1.3 and 1.8
• Agents with an HR >1.8 are not approvable

The Rate of Dying From CVD Has Dropped Dramatically in Diabetes^[24]

• A dramatic reduction in the risk of dying from CVD events occurred between 1998 and 2004
• This risk has fallen from a 2.4-fold higher risk in 1998 to a 2.0-fold higher risk in 2004

Number (in Millions) of Persons With Diabetes Aged ≥35 With Self-Reported Heart Disease or Stroke^[25]

• The increased number of patients with diabetes who report CVD or stroke over the last decade is due to the increased number of patients with diabetes

Metabolic Components of Diabetes: ADA Treatment Recommendations ^[12,26-28]

• Multifactorial treatment of CV risk factors, especially cholesterol, is important

Summary

Examining the Data Across T2D Treatment: Older Agents

UKPDS Study: Fatal or Nonfatal MI, Sudden Death^[27,29]

• In the UKPDS study, intensive blood glucose lowering just missed statistical significance in lowering risk for nonfatal MI or sudden death
• Only 342 overweight individuals in the metformin group had statistically significant reductions in adverse events

UKPDS Study: Metformin^[29]

• In the UKPDS study, metformin showed more beneficial effect than sulfonylureas or insulin for
  • Any diabetes endpoint
  • All-cause mortality
  • Stroke
• A UKPDS substudy of sulfonylurea monotherapy vs sulfonylurea + metformin showed an increased risk of diabetes-related death in the sulfonylurea + metformin group, probably a statistical fluke
• Metformin may be first-line pharmacotherapy for patients who are overweight and have diabetes

10-Year Follow-Up of UKPDS: Mean HbA_1c and Body Weight^[30]

• Patients were followed observationally
• In the intervention group, HbA_1c was hard to sustain in usual care
• Conventional group saw reduction in HbA_1c
• For most of the 10-year follow-up, HbA_1c and weight were similar in the metformin and sulfonylurea-insulin groups

10-Year Follow-Up of UKPDS: HR for Prespecified Aggregate Clinical Outcomes^[30]

• For any diabetes endpoint in the 10-year follow-up of the UKPDS study
  • Both intensive treatment groups did better
  • Benefits were statistically significant in both groups after study
• For MI
  • The sulfonylurea-insulin group missed statistical significance for HR
  • Benefits were statistically significant in metformin group

10-Year Follow-Up of UKPDS: HR for Prespecified Aggregate Clinical Outcomes (cont)^[30]

• Reductions in microvascular disease through the follow-up period
  • Remained statistically significant in the sulfonylurea-insulin group
  • Were not statistically significant in the metformin group, probably a function of small numbers in that study
• Reductions in death from any cause
  • Were statistically significant in the sulfonylurea-insulin group
  • Remained statistically significant in the metformin group

UKPDS: Effect of Earlier Glucose Control^[27,30]

• Any diabetes-related endpoint, microvascular disease, MI, and all-cause mortality all improved in the intensive treatment group
• Through most of follow-up period, the intensive treatment group had similar HbA_1c levels as the conventional group
• "Legacy effect" of early better control was carried long-term

Metformin Use Among Patients With T2D and Atherothrombosis^[31]

• Patients were followed for 2 years to assess the effects of metformin use on mortality
• All-cause mortality was 25% lower among patients who used metformin
• Metformin may decrease mortality when used as secondary prevention

Comparative Effectiveness of Sulfonylurea or Metformin Monotherapy on CV Events in T2D^[32]

• The effects of sulfonylureas and metformin on CV outcomes in patients with T2D were examined in a study of data from linked VA and Medicare databases
• The metformin group had a lower incidence of CV events than the sulfonylurea group

Sulfonylureas and CV Risk^[33]

• The CV safety of sulfonylureas was examined in a meta-analysis
• The incidence of major adverse cardiovascular events (MACE, 29 trials) was equivalent in the sulfonylurea and comparator groups
• CV mortality (37 trials) was higher for sulfonylureas vs comparators

Metformin + Sulfonylurea Combination Therapy Is Associated With Increased Mortality Risks^[34]

• A meta-analysis of 9 trials was conducted
• Metformin plus sulfonylurea increased the relative risk of a composite endpoint of
  • CV hospitalization/mortality
  • Fatal/nonfatal events
• Effect was irrespective of reference group (diet, metformin, or sulfonylurea monotherapy)
• Meta-analyses and database analyses are lower level of evidence than prospective randomized controlled trials (RCTs)

Ongoing CAROLINA^® Trial^[35]

• The ongoing CAROLINA^® trial is examining linagliptin (a dipeptidyl-peptidase-4 [DPP-4] inhibitor) vs a sulfonylurea
• More than 6000 patients have been enrolled
• Results are expected in September 2018

PROspective Actos Clinical Trial in Macrovascular Events Study Design^[36]

PROactive Primary Endpoint^[36]

• A 10% nonsignificant reduction in the primary endpoint was seen with pioglitazone
• HbA_1c was lower in the pioglitazone group

PROactive Prioritized Secondary Endpoint: Composite of All-Cause Mortality, Nonfatal MI, and Stroke^[36]

• A 16% reduction in the secondary endpoint was statistically significant

PROactive: Pioglitazone (Pio) Reduces "Hard" Coronary Heart Disease Endpoints^[37]

• In a post hoc analysis of data from PROactive on patients with previous MI, the effects of pioglitazone on recurrent MI were examined
• Occurrence of fatal or nonfatal MI (excluding silent MI) was reduced by 28% with pioglitazone
• Occurrence of ACS was reduced by 37%

PROactive: Time to Fatal/Nonfatal Stroke in Patients With Previous Stroke^[38]

• In a post hoc analysis, a 47% reduction in time to fatal/nonfatal stroke was seen among patients in the pioglitazone group with previous stroke
• No difference was observed among patients without a previous stroke

PROactive Subgroup Analyses^[37,38]

PROactive: HF Hospitalization and Mortality^[36]

• The significant increase in risk of hospitalization for HF seen in the PROactive study is sometimes seen with thiazolidinediones (TZDs)
• No statistically significant increase in risk of fatal HF was observed

Rosiglitazone: Risk of MI and Death^[21]

Rosiglitazone Evaluated for CV Outcomes in Oral Agent Combination Therapy for T2D (RECORD): Time to CV Death or CV Hospitalization^[39]

• No increase in CV death or CV hospitalization was observed in the RECORD study

RECORD: Deaths and Hospitalization From CV Causes^[39]

• The randomized RECORD study provides a stronger level of evidence than the 2007 meta-analysis
• No statistically significant increase was found in the relative risk for CV death, all-cause death, MI, or stroke between the rosiglitazone and active control groups
• An increased risk of HF was observed in the rosiglitazone group

BARI 2D: Rosiglitazone and Mortality and CV Endpoints^[40]

• A reduction in the risk of stroke with rosiglitazone treatment was seen in the BARI 2D study

IRIS Trial: Post-Stroke EFS of Pio in Patients With Insulin Resistance^[41]

• The IRIS trial enrolled patients who had prior stroke and insulin resistance but no diabetes
• Pioglitazone reduced stroke or MI by 24% vs placebo
• A reduction in patients progressing from insulin resistance to actual T2D was observed

STOP-NIDDM Trial: Acarbose Prevents T2D and CV Events in Patients With Impaired Glucose Tolerance^[42]

• α-glucosidase inhibitors are not commonly used in the United States
• In the STOP-NIDDM study, a 50% reduction in CV events was observed in patients receiving acarbose
• Postprandial hyperglycemic was not measured, only presumed
• Methodological issues in this study include a potential statistical fluke

Acarbose CV Evaluation

• Examining whether acarbose reduces future events and/or progression of diabetes in patients with impaired glucose tolerance who have already had CV events
• Trial is nearing conclusion

The ORIGIN Trial

• Insulin was the first antihyperglycemic agent

ORIGIN: Methods^[43]

ORIGIN: MI, Stroke, or Death From CV Causes^[43]

• No difference in CV outcomes was found between standard care and insulin glargine treatment
• The ORIGIN trial met the FDA safety standard, although it was not part of the CV outcomes trials

ORIGIN: HbA_1c and Other Outcomes^[43]

• Early initiation of treatment resulted in low HbA_1c at end of 7 years in both groups
• The insulin group had more severe cases of hypoglycemia, but the number of events was low
• Weight gain with insulin was statistically significant but not huge

ORIGIN: Cancer and New T2D^[43]

• No increased risk of cancer was seen in the ORIGIN study

ORIGIN: Conclusions^[43]

Summary

Ochsner Medical Center New Orleans, LA

Examining the Data Across T2D Treatment: Modern Agents

DPP-4-I: Alogliptin, Linagliptin, Saxagliptin, Sitagliptin^[44]

• Food is absorbed from the intestine
• Glucagon-like peptide-1 (GLP-1) stimulates insulin secretions and suppresses glucagon secretion
• Hypogluconemia is a significant component of T2D
• Blocking GLP-1 leads to a 2-fold increase in native GLP-1 in the blood stream

SAVOR-TIMI, EXAMINE, and TECOS

SAVOR-TIMI, EXAMINE, and TECOS: Baseline Characteristics ^[45-48]

SAVOR, EXAMINE, and TECOS: Key Results

• In the SAVOR study (saxagliptin vs placebo), a surprisingly higher incidence of hospitalization for HF was seen in the saxagliptin group^[45]
• The EXAMINE study (alogliptin vs placebo) had a short median follow-up^[46]
• In the TECOS study (sitagliptin vs placebo), no signal for HF was observed^[48]

SAVOR-TIMI: Risk of HHF With Saxagliptin vs Placebo^[49]

• People with an estimated glomerular filtration rate (eGFR) ≤60 had higher risk of hospitalization for HF (HHF)
• People with a history of HF had a higher risk of HHF

FDA Exploratory Review of EXAMINE: Time to HHF^[50]

FDA Drug Safety Communication DPP-4-I, April 5, 2016^[51]

Ongoing CAROLINA^® and CARMELINA Trials^[35,52]

SGLT2 Inhibitors: Canagliflozin, Dapaglifozin, Empagliflozin ^[53-55]

• Glucose is reabsorbed in the proximal convoluted tubule of the kidney after filtration
• Leveraging this mechanism leads to glycosuria
  • Reduction in plasma glucose
  • Reduction in HbA_1c
  • Osmotic diuresis
  • Caloric loss
  • Weight reduction

EMPA-REG OUTCOME

EMPA-REG: Study Design^[56]

• Patients were randomized 2:1 to empagliflozin or placebo

EMPA-REG: Baseline Characteristics -- T2D^[56]

• Patients were at high risk for CV events

EMPA-REG: Baseline Characteristics – CVD^[56]

EMPA-REG: Primary Outcome 3-Point MACE^[56]

• First time an antihyperglycemic agent demonstrated reduction in CV risk

EMPA-REG: CV Death^[56]

EMPA-REG: CV Death, MI, and Stroke^[56]

• The HR for the primary outcome of death from CV causes, nonfatal MI, or nonfatal stroke was 0.86, and was highly statistically significant
• Rates of nonfatal MI did not significantly differ between groups
• Rates of nonfatal stroke favored placebo, but the difference was not statistically significant

EMPA-REG: Hospitalization for HF^[56]

• A 35% reduction in hospitalization for HF was observed in the EMPA-REG study

EMPA-REG: Key Results^[56]

Empagliflozin to Be Studied for the Treatment of People With Chronic HF^[57]

Ongoing SGLT2 Inhibitor CV Outcomes Studies^[58-60]

GLP-1 RAs: Albiglutide, Dulaglutide, Exenatide (BID, QW), Liraglutide, Lixisenatide,* Semaglutide*^[61]

CV Effects of GLP-1 Receptor Activation^[62]

• GLP-1 receptor activation produces a pharmacologic level of GLP-1 in the bloodstream
• DPP-4 inhibition offers physiologic doubling, but a pharmacologic level is needed for weight reduction

ELIXA

ELIXA Study Design^[63]

• The ELIXA study enrolled an "extraordinarily sick group of people"
• Lixisenatide is a short-acting once-GLP-1 receptor agonist (RA)

ELIXA: Key Study Population and Baseline Characteristics^[63]

ELIXA: ITT Analysis^[63]

• Lixisenatide met criteria for noninferiority but not superiority to placebo

ELIXA: Mean HbA_1c Level, According to Study Visit^[63]

• A difference of−0.27 percentage points (95% CI, −0.31 to −0.22 in mean HbA_1c was maintained throughout the trial in the lixisenatide group

LEADER

LEADER: Primary Objective^[64]

LEADER: Patient Population^[64]

LEADER: Patient Characteristics^[64]

LEADER: Baseline Characteristics^[64]

• HbA_1c was higher than in other outcomes trials

LEADER: What We Know So Far, Topline Results

FREEDOM-CVO Trial: A Study to Evaluate CV Outcomes in Patients With T2D Treated With ITCA 650^[65]

Ongoing GLP-1 RA CV Outcomes Studies^[66-68]

Practical Implications: What Does It Mean for My Practice and Patients

Patient Case 1

Polling Question: What Would You Add Next? Audience Response

Patient Case 2

Polling Question: What Would You Add Next? Audience Response

Question and Answer

Question From the Audience: Does the panel think SGLT2 inhibitors should be distinguished as the second-line class after metformin based on CV outcomes trial data?

• Dr Anderson
  • The CV outcome trial benefit is in a unique group of individuals
  • The GLP-1 RA class has good benefits in terms of weight reduction, nonglycemic benefits, and probably better efficacy
• Dr Jabour
  • We do not know how the EMPA-REG findings apply to younger patients without CVD
  • LEADER is positive, ELIXA is neutral. We do not yet know the best second-line agent
• Dr Wysham
  • I would choose empagliflozin for a patient with heart disease

Question From the Audience: How have the CV outcome trials affected guidelines so far, and how do you think they will impact them in the future?

• Dr Blonde
  • After metformin, therapy should be individualized, and the order should be GLP-1 RA first and SGLT2 inhibitor second
  • Canadian guidelines changed the order for people with CVD. Patients similar to the patients in the EMPA-REG trials should receive an SGLT2 inhibitor with proven benefit in reducing CV events

Question From the Audience: What does the panel think CAROLINA^® results may do to guidelines? Particularly since it is a study of a DPP-4 inhibitor vs a sulfonylurea?

• Dr Blonde
  • The American Diabetes Association (ADA) and European Association for the Study of Diabetes (EASD) list 6 choices, including sulfonylureas, to add to metformin, and the choice is individualized
  • The American College of Endocrinology (ACE) algorithm emphasizes treatments with low risk for hypoglycemia
  • CAROLINA^® is not likely to affect the guidelines
• Dr. Anderson
  • Most sulfonylurea use is done reluctantly for reasons of cost
  • Hopefully better agents will be made available for patients

Question From the Audience: Today, Dr Ele Ferrannini presented in an oral session, and concurrently published in Diabetes Care, the hypothesis that β-hydroxbutyrate is taken up by the heart and oxidized in preference to fatty acids, and contributed to the effects seen with empagliflozin in EMPA-REG....what is the panel's thought?

• Dr Blonde
  • Both are related to fuel energetics potentially having a role
  • It makes the heart more efficient at making ATP with lower oxygen demands
• It is compelling for the kidney as well as the heart

Summary and Closing Remarks

Large Non-Insulin CV Outcomes Trials in T2D