September 07, 2016
19 min read

Beyond safety, new diabetes drugs offer CV benefit

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

The medical community has long recognized the strong relationship between CVD and diabetes. Compounding concerns over this association have been questions about the CV safety of diabetes treatments, in particular the effect of these medications on the risk for CV mortality, MI, stroke and HF hospitalization.

Cardiology Today assembled a distinguished panel of experts to discuss the topic of CV effects of diabetes medications and what these effects might mean for treating patients. A particular focus was on newer drugs.

In the wake of an FDA requirement that all new diabetes drug trials must further investigate CV risk, several agents have been documented to pose no elevated risk. Recent investigations, however, have revealed unanticipated benefit in terms of risk reductions for certain, but not all, glucagon-like peptide-1 (GLP-1) receptor agonists and at least one sodium-glucose cotransporter-2 (SGLT2) inhibitor, the latter of which is under FDA consideration for an indication for prevention of CV death and hospitalization for HF.

This expert round table tackles topics including the CV effects of different diabetes therapies, possible mechanisms of action, recent trial results and more. Read on for insight from some of the leading experts in this area.

Diabetes and CV risk

Carl Pepine, MD, MACC: This is clearly a critical topic for those of us who care for patients with, or at high risk for CVD. Keith, will you begin our discussion?

Keith C. Ferdinand, MD: It is important to understand that the reason why we are having this conversation is that, while diabetes may not be a CAD risk equivalent, patients with diabetes clearly die of heart disease more than any other cause. If you look across the population with diabetes, approximately 50% will develop HF and more than three-quarters will have elevated BP or are on anti-hypertensive medication. Within the multicultural U.S. population, specific racial/ethnic groups such as South Asians, blacks and certain Hispanics, especially Mexican Americans, have the highest rates of diabetes.

George L. Bakris, MD: I recently saw an interesting statistic that one in five Americans are projected to be obese by 2050. Obesity is a precursor to diabetes and heart disease. Further, if you follow that to its logical conclusion and assume that 30% to 35% of all people with diabetes are going to develop kidney disease, then we will have skyrocketing CV events. Because, while diabetes alone is not an MI equivalent, the combination of kidney disease and diabetes is well beyond an MI equivalent in terms of increased risk for adverse CV outcomes.

There are a number of new oral agents that have been around for a while and now clinical trial data are coming out about them. The SGLT2 inhibitors and the GLP-1 receptor agonists definitely are showing clear CV benefit, not only on surrogates like BP and glycemic control, but also on actual adverse outcomes. These therapies have limitations in their use, based on the patient’s kidney function, but are very promising agents and, importantly, neither class is associated with hypoglycemia.

Round Table Participants


Pepine: Let’s discuss results with the first large outcome trial we have with the SGLT2 inhibitors, EMPA-REG OUTCOME. Data show that empagliflozin (Jardiance, Boehringer Ingelheim) was associated with decreased risk for HF hospitalization, CV death and other HF-related endpoints, regardless of whether patients had HF at baseline.

Darren K. McGuire, MD, MHSc: SGLT2 is a cotransporter in the kidney in the proximal tubule epithelium; it is responsible for immediate reclamation of filtered glucose. Inhibiting SGLT2 leaves glucose in the urine and causes urinary glucose excretion. It resets the threshold of serum glucose at which excretion in the urine will occur, so once circulating glucose gets to or below that level, there’s no further excretion and almost no incremental risk for hypoglycemia. Patients will excrete between 60 g to 100 g of glucose a day, which translates to loss of up to 400 calories a day. So, the SGLT2 inhibitors, over long periods of time, are associated with robust and sustained weight reduction.

Darren K. McGuire, MD, MHSc, from University of Texas Southwestern Medical Center, and other experts discuss the role of cardiologists in the management of diabetes.
Darren K. McGuire, MD, MHSc, from University of Texas Southwestern Medical Center, and other experts discuss the role of cardiologists in the management of diabetes.

Source: Robert Levy Photography.

EMPA-REG OUTCOME was a double blind trial of 7,020 patients with prevalent CVD and type 2 diabetes who were randomly assigned to treatment with empagliflozin 10 mg or 25 mg or placebo. It focused on a three-point major adverse CV event outcome, a standard CV outcome for clinical trials in cardiology. For the first time, a large-scale dedicated CV outcomes trial of an antihyperglycemic therapy showed a significant improvement in outcomes, a 14% risk reduction for the composite outcome of CV death, MI and stroke. Importantly, the outcome was driven almost completely by CV death reduction. We’ve never seen anything like this, not only in the diabetes space but in the CV space — a 32% risk reduction for mortality.

Pepine: The mortality curves separated very early within the first year.

McGuire: There is a lot of speculation about why that occurred. It’s way too early for glucose control and way too early and too large a magnitude for it to be attributable to BP. It’s not attributable to the osmotic diuresis (or the natriuresis, as sodium reclamation is also inhibited by blocking SGLT2). There’s something more complex going on, likely at the level of the kidney; much recent speculation has to do with acute alteration of renal hemodynamics.

Prakash Deedwania, MD: An important point is that there was a substantial reduction in HF-related events and, fortunately, you can say it’s not osmotic diuresis. But we also know that most diabetic patients have diastolic dysfunction vis-à-vis preserved HF or preserved ejection fraction (HFpEF). My speculation is that by stabilizing that and by having this weight loss and the diuresis that perhaps the progression of HFpEF is significantly decreased and there could be beneficial effects of neurohormones, which eventually tie in with the subsequent outcome of mortality.


Ferdinand: Anything we say now is speculation because we have not been able to find the mechanism for that early separation in death. There may have been more people with HF than we know. The researchers did not measure brain natriuretic peptide (BNP) levels, and 10% of the patients on the questionnaire noted a previous history of HF, 70% had MI or stroke; the mean age was 63 years.

Pepine: Are we expecting that the other SGLT2 inhibitors are going to do the same thing?

McGuire: Yes, I am. Of course, we’ll only know that when we get the data so I’m not ready to say we can conclude that now. An interesting thing that happened here is we saw these same point estimate plots and CV outcomes curves with the preliminary data for dapagliflozin (Farxiga, AstraZeneca) and canagliflozin (Invokana, Janssen) that supported their FDA New Drug Applications. When we saw the mortality estimate going in the right direction early on with only 40 or so events, we thought it was just too good to be true with imprecision of effect estimates based on such small numbers. It’s interesting going back historically because we completely dismissed all of these signals early on.

The other intriguing point is that there is a persistent stroke signal. With BP reduction and diuresis and all these things that go with it, you would expect, especially over a longer period of time, to have some stroke reduction, or at least a neutral effect. The point estimates, even in EMPA-REG OUTCOME, were slightly in the wrong direction. Now, it’s not statistically significant and we cannot draw conclusions from that. But we see in the dapagliflozin phase 2/phase 3a data a point estimate of neutrality despite the BP effects and all the other beneficial side effects. With canagliflozin, it was the biggest concern of the entire New Drug Application that stroke went in the opposite direction from the composite MACE outcome and with a similar magnitude. Again, these are small numbers and, hopefully, over time this will play out to neutrality or benefit, but what we can see now is some discordance and this doesn’t appear to be atherosclerotic vascular disease mitigation.

Other classes of interest

Pepine: Let’s discuss a different class of antidiabetes therapy: the GLP-1 receptor agonists. Liraglutide (Victoza, Novo Nordisk) is approved for type 2 diabetes and lowers glucose levels. Recent results of the LEADER trial found reduced risk for CV death, nonfatal MI and stroke with liraglutide vs. placebo in patients with type 2 diabetes and a history of CVD. During a mean follow-up of 3.8 years, liraglutide reduced risk for major adverse cardiac events by 13%, all-cause death by 15% and CV death by 22% vs. placebo, while reducing both HbA1c and body weight.

Its use has been associated with reductions in weight and BP, but it has been shown to increase heart rate.

Darren, can you comment on other GLP-1 receptor agonists?

McGuire: The GLP-1 receptor agonist lixisenatide (Lyxumia, Adlyxin; Sanofi) was studied in the ELIXA trial. Patients were enrolled within 30 to 180 days after an ACS and randomly assigned to treatment with lixisenatide or placebo. The data demonstrated a completely neutral result, a robust noninferiority result, thus substantiating its presence on the market and ultimate FDA approval for CV safety. There was no signal of incremental efficacy.

Carl J. Pepine, MD, MACC, (right) leads a discussion of the results of the EMPA-REG OUTCOME trial.
Carl J. Pepine, MD, MACC, (right) leads a discussion of the results of the EMPA-REG OUTCOME trial.

All images: Robert Levy Photography.

Ferdinand: Another once-weekly GLP-1 receptor agonist available in the United States is dulaglutide (Trulicity, Eli Lilly). Data we published in Hypertension in 2014 showed BP lowering at 1.5 mg but an increase in heart rate of 3 bpm to 4 bpm vs. placebo. This was an international study of 755 patients that showed significant BP lowering on 24-hour ambulatory BP monitoring. Also, we published this year a retrospective analysis of phase 2 and phase 3 trials of dulaglutide and the data showed similar CV safety and potential benefits.

Overall, these effects may be mediated by direct actions of GLP-1 receptor agonists on the heart, sympathetic nervous system, or both mechanisms. In the LEADER trial, heart rate was 3 bpm (95% CI, 2.5-3.4) higher with liraglutide vs. placebo, but it did not seem to have deleterious effects. More research is needed on heart rate, but GLP-1 and its degradation products may cause vasodilation through GLP-1 receptor-dependent and -independent pathways that lower BP and increase heart rate.

We await the final outcome studies of FDA-approved GLP-1 receptor agonists. A small increase in heart rate is something we should pay more attention to. So, perhaps in a patient without HF, the GLP-1 receptor agonists would be beneficial, but more data are needed to clarify the heart rate concerns.

George L. Bakris, MD, debates the best course of medical management for patients with diabetes and kidney disease.
George L. Bakris, MD, debates the best course of medical management for patients with diabetes and kidney disease.

Bakris: There have been some interesting small studies with GLP-1 receptor agonists and SGLT2 inhibitors that look at BP, and it’s pretty much additive.

Deedwania: Another new molecule that we don’t know too much about is the first PPAR alpha/gamma agonist — saroglitazar — that we learned is safe. We have had two PPAR alpha/gamma agonists — muraglitazar and aleglitazar — thus far, which were not safe. According to recent data, 1-year experience with saroglitazar showed no renal or hepatic adverse effects and reductions in HbA1c of 0.8% to 1%. In addition, it does what PPAR-alpha agonists do: reduce triglycerides and increase HDL. We need to have longer experience and we need randomized clinical trials to evaluate long-term safety and impact on clinical outcomes.

Ferdinand: That would be interesting. We have seen this before, especially with antidiabetes medications, where they look beneficial early on in data but actually have adverse effects later.

McGuire: We also have the dipeptidyl-peptidase IV (DPP-IV) inhibitors. These are once-daily tablets that inhibit the endogenous degradation of GLP-1. By inhibiting its degradation, these agents potentiate the endogenous effects. There is almost no hypoglycemia and again no associated weight gain, so they are very popular in the diabetes community.


There are now three randomized clinical trials completed showing unequivocal CV safety from the three-point major adverse CV events endpoint with saxagliptin (Onglyza, AstraZeneca), sitagliptin (Januvia, Merck) and alogliptin (Nesina, Takeda). There’s a fourth drug available, linagliptin (Tradjenta, Boehringer Ingelheim). Linagliptin is unique in the class in that it’s not renally excreted, so it’s the only one that may be effective in patients with kidney dysfunction. The CV outcomes studies for linagliptin are ongoing, one compared with the sulfonylurea glimepiride and the other compared with placebo.

The caveat is that saxagliptin looks different than the others in that it has a significant increased risk for hospitalization for HF. We have no idea what that mechanism might be, but it’s a compelling finding from the very large randomized SAVOR-TIMI 53 trial. A similar trend for increased HF was observed with alogliptin in the EXAMINE trial. Based on these observations, the FDA recently required product label modification warning about the potential HF risk with these two medications; no such HF signal was observed with sitagliptin in the TECOS trial, and that drug was spared the product label modification regarding HF.

Ferdinand: The New England Journal of Medicine recently published a meta-analysis of 1.5 million patients on incretin-based drugs and hospitalization for HF. In this analysis, incretin-based drugs were not associated with an increased risk for hospitalization for HF compared with commonly used combinations of oral antidiabetic drugs.

McGuire: We should also discuss use of pioglitazone: an old drug with a new pair of shoes. Pioglitazone was dragged down inappropriately with the rosiglitazone debacle. I’ve always been an advocate of pioglitazone as a second-line drug behind metformin. Pioglitazone is generic now.

We recently saw results of the NIH-funded IRIS trial, which included patients with a recent stroke or transient ischemic attack who had insulin resistance fairly rigorously defined, but not with diabetes. They were randomly assigned to pioglitazone or placebo. Results, ultimately, showed a 24% risk reduction in three-point major adverse CV events. This was a favorable and clearly statistically superior trial of pioglitazone. It is superimposable on the PROACTIVE results. If you look at the three-point major adverse CV events outcome, that had about a 20% RR reduction.

Bakris: I took a step back because, as much as I like pioglitazone, the volume is dose-dependent. The second issue is bone fractures.

McGuire: Another serious, but uncommon, complication is macular edema.

Table. Results of large non-insulin CV outcomes trials in patients with type 2 diabetes

Guidance for cardiologists

Pepine: My perception is that cardiologists are being called upon more and more to manage diabetes. What recommendations do you have for our readers that would help guide them in terms of what drugs to use?

Prakash Deedwania, MD, and Darren K. McGuire, MD, MHSc, discuss the role of metformin in patients with diabetes at risk for CVD.
Prakash Deedwania, MD, and Darren K. McGuire, MD, MHSc, discuss the role of metformin in patients with diabetes at risk for CVD.

Deedwania: Both metformin and insulin are available. Because many patients with diabetes have chronic kidney disease (CKD), metformin has to be used with significant caution. Insulin also increases the risk for hypoglycemia in patients with diabetes and CKD. The caution would be that newer drugs may be the safest because they don’t have as much of the relationship to hypoglycemic issues we see with metformin.

Ferdinand: This is an important concept. Is the contraindication of metformin with CKD overstated because the lactic acidosis is quite rare?

Bakris: There is currently a debate about this: When do you cut off metformin? When can you start it? In my opinion, it is appropriate to use maximal-dose metformin if the estimated glomerular filtration rate (eGFR) is above 60 mL/min/1.73 m2. Period. End of discussion. If the eGFR is between 45 mL/min/1.73 m2 and 60 mL/min/1.73 m2, it’s appropriate to start metformin at a half dose, 500 mg twice daily. If the eGFR is below 45 mL/min/1.73 m2, should you start metformin? That issue is controversial. Most people would say no. But if a patient is on metformin and doing well, I would keep them on metformin. But, if the eGFR is below 30 mL/min/1.73 m2, the party is over. There is a new metformin formulation that generates much lower lactate levels and may be useful in people with eGFR down to 30mL/min/1.73 m2 regardless of what eGFR it is started; however, this remains to be seen.

McGuire: These cautions are only in the United States. In the United Kingdom and most other places that have guidance, metformin can be used now safely to an eGFR of 30 mL/min/1.73 m2. Between 45 mL/min/1.73 m2 and 30 mL/min/1.73 m2 or 60 mL/min/1.73 m2 and 30 mL/min/1.73 m2 or depending on which guideline you read they will reduce dose by half. There’s not a single case of clearly metformin-associated lactic acidosis in the comparative data set of a drug that we’ve had available in the world for 60 years now.

Ferdinand: I’m in favor of loosening controls on use of metformin. However, we should use caution in the person who is NPO for cardiac intervention and has not been given IV hydration and was on metformin which was not discontinued and then has a high angiographic contrast load.

Bakris: Let me give you a reference. There’s a study on the use of diabetes drugs in people with advanced CKD, with eGFR below 30 mL/min/1.73 m2 and/or on dialysis. There is a very limited repertoire for this group. In fact, the SGLT2 inhibitors by and large go out the window. Of the GLP-1 receptor agonists, liraglutide is the only agent that you can really do anything with in this group. Of the sulfonylureas, only glipizide can be used. It’s a very small list in that group. Most of these patients are on ACE inhibitors or angiotensin receptor blockers. If they’ve got an eGFR in the 30s or 40s and they develop diarrhea for a few days, they can develop acute kidney injury. I’ve seen this reported in trials.

Pepine: On April 8, 2016, the FDA revised warnings regarding use of metformin in certain patients with reduced kidney function. The FDA concluded from a review of published studies that metformin can be used safely in patients with mild impairment in kidney function and even in some patients with moderate impairment in kidney function. The agency has required changes to the metformin labeling to reflect this new information and provide specific recommendations of metformin’s use in patients with mild to moderate kidney impairment. Further, the FDA also recommends that the measure of kidney function used to determine whether a patient can receive metformin be changed from one based on a single laboratory parameter (blood creatinine concentration) to one that provides a better estimate of kidney function in patients with kidney disease (ie, eGFR).

McGuire: Coming back to the SGLT2 inhibitors and kidney disease, a very intriguing observation across the class is a paradoxical interaction of BP reduction by eGFR, with an inverse association of greater BP reduction with lower eGFR even after adjustment for baseline BP. By the mechanism of action, if it’s just diuresis causing the BP reduction, as eGFR wanes so should the BP reduction, but it’s the opposite. So this is a vasoactive medication, or at least it has indirect vasoactive effects.


Diabetes therapy of choice

Pepine: If you had your choice of the drugs that we have been talking about today, what would you use?

McGuire: For patients with or at high risk for CVD, I favor empagiflozin and liraglutide, based simply on the CV outcomes data on the backdrop of excellent safety and tolerability. However, this is with the key caveat that these newer antihyperglycemic therapies are very expensive. If we can get a way for our patients to get them paid for or pay for them, I believe they represent a good option.

Deedwania: I would go with a DPP-IV inhibitor to start.

Ferdinand: I would start metformin and add a GLP-1 receptor agonist.

McGuire: I am also still using metformin. Those are pretty good options. After that, you’re just making guesses.

Managing HbA1c, BP

Pepine: What’s your recommended HbA1c goal for these patients?

Bakris: If I can get them to 7% or slightly lower, I’m ecstatic. I don’t shoot for anything lower.

McGuire: I’m happy at 8% for most patients. For example, if it is a young, otherwise healthy, 50-year-old with single-vessel CAD and preserved left ventricular function, down to 7% would be appropriate, but I wouldn’t push below 7%. For most CV patients who are older and with more comorbidities, the American Diabetes Association/European Association for the Study of Diabetes guidance since 2012 has endorsed a target of 8% or possibly lower for such patients with type 2 diabetes.

Deedwania: I try for somewhere between 7% and 7.5%.

Ferdinand: For middle-aged patients, I’m happy with a goal of 7.5%. With 8%, I’m a bit uncomfortable. I think there’s a microvascular penalty for hyperglycemia and maybe a macrovascular penalty, although it is delayed by a decade or more. The microvascular disease is not delayed and does not just mean kidney disease; it also means retinopathy and even maybe limb loss.

Keith C. Ferdinand, MD, says patients with diabetes are at extremely high risk for CVD.
Keith C. Ferdinand, MD, says patients with diabetes are at extremely high risk for CVD.

Pepine: So, considering interactions between BP and diabetes, what are you comfortable with now that we have these new therapies?

Ferdinand: Increased age is the most powerful predictor of risk, next is systolic BP. There’s no reason to believe that systolic BP should be over 150 mm Hg for a person aged 60 years or older because we didn’t have rigorous evidence based on clinical trials that giving drugs to a goal less than 140 mm Hg was beneficial. The SPRINT trial did not include persons with diabetes, but there are several reasons why SPRINT showed lower was better in persons without diabetes, which is what the ACCORD trial did not show. One is sample size: SPRINT was twice as big — 9,000 participants vs. 4,700 participants. Second, SPRINT participants were older and had more disease. I believe that 140 mm Hg is a reasonable goal for a person with diabetes. But, if tolerated, I would rather have that person with a systolic BP less than 130 mm Hg; if they’re younger, even lower.

Pepine: The median fasting blood sugar in SPRINT at entry was 99 mg/dL, so half of them had prediabetes. As they said very clearly in their discussion, there’s no reason to believe that the patients in SPRINT behaved any different than those with frank diabetes.

Also important for this discussion is the more recent ACCORD BP trial data, concluding that intensive BP (< 120 mm Hg) or intensive glycemic treatment alone improved major CVD outcomes without additional benefit from combining the two.

So what BP target are we going to recommend here?

Carl J. Pepine, MD, MACC, notes that cardiologists are increasingly being asked to manage diabetes.
Carl J. Pepine, MD, MACC, notes that cardiologists are increasingly being asked to manage diabetes.

Deedwania: Somewhere in between. Most recently, there was a paper that came out of 700,000 patients with meta-regression analysis of all the data to date showing clearly that what Keith is saying: BP less than 150 mm Hg is quite beneficial; 140 mm Hg is the threshold at which they found reduction in every event related to hypertension in these diabetic patients. If you go to 130 mm Hg you start seeing the hazard ratio go on the other side. I think there’s a sweet spot between 140 mm Hg and 130 mm Hg. I think the speculation is 135 mm Hg may be appropriate.

Having said that, there are patients with diabetes, and we are seeing an increasing number of these patients who are younger, have shorter duration of diabetes, and if they have already shown albuminuria or other manifestations, perhaps it’s reasonable. We have no evidence one way or the other, that it is OK to reduce their systolic BP to somewhere close to 130 mm Hg.

Ferdinand: We’re not going to get definitive evidence one way or another. In order to treat the patient optimally, in that particular case, we have to be what’s called a “good doctor.”

Bakris: I also think a range for BP is reasonable. We’re not physicists; we’re biologists. People have variability of BP all the time. I think if you want a sweet spot I would say the range of BP somewhere between 128 mm Hg on the low end and 133 mm Hg on the high end, somewhere in there is where we should be. If you convert the SPRINT data to the way office BP is measured routinely, this is the range you will get.


Future directions

Pepine: What is your take-home message?

McGuire: With the EMPA-REG OUTCOME trial data, we now have a drug for antihyperglycemic therapy that mimics lipid-lowering therapy with statins. It’s a single dose; it’s once-daily; it doesn’t need to be titrated; and a cardiologist can prescribe it. The interpretation of the data I’ve seen is we have a CV medication that has a side effect of glucose lowering. So I hope we don’t consider this just a diabetes drug, although that’s the only population it’s approved in.

Deedwania: We have to recognize that diabetes is becoming an epidemic all over the world. We now have newer drugs that have fewer side effects, are better tolerated, and clearly that calls for us to intervene in the diabetic sphere early on in terms of whether it’s glucose control, BP control, and certainly the most beneficial thing that we are talking about in diabetic patients is use of statins. The statins really have a very positive effect.

Ferdinand: In adult patients with CVD, middle-aged and older, the rates of diabetes are much higher than in the general population, maybe as high as 30%. There are not enough endocrinologists to treat the large and growing number of people with diabetes. We, as cardiologists, are going to have to step forward, especially in those patients who have concomitant CVD, and treat their glucose.

Bakris: We need patient education so that patients understand why they’re taking multiple pills. Combination pills are huge, and something like an SGLT2 inhibitor combined with metformin, which already exists, is really good. BP goals as we’ve defined them, I fully agree with Prakash, statins clearly have a role here. With glycemic control, that trifecta will reduce not only mortality but CV benefits and HF. – Compiled by Katie Kalvaitis

Disclosure: Bakris is principal investigator of FIDELIO, a Bayer-funded renal outcome study, and reports consulting for AbbVie, Bayer, Boehringer Ingelheim, Janssen, Medtronic, NxStage, Relypsa, Takeda and Vascular Dynamics. Deedwania reports consulting and/or speaking on behalf of Amgen, Novartis, Pfizer and Sanofi. Ferdinand reports consulting for Amgen, Boehringer Ingelheim, Eli Lilly, Quantum Genetics and Sanofi, and conducting research for Boehringer Ingelheim. McGuire reports receiving personal fees for clinical trial leadership from AstraZeneca, Boehringer Ingelheim, Eisai, GlaxoSmithKline, Janssen Research and Development LLC, Lexicon, Lilly USA, Merck Sharp and Dohme, Novo Nordisk and Takeda Pharmaceuticals and reports receiving personal consultant fees from Merck Sharp and Dohme, Novo Nordisk, Regeneron and Sanofi Aventis Group. Pepine reports no relevant financial disclosures.