What Cardiologists Need to Know About New Diabetes Drugs

An Interview With Dr Darren McGuire

Interviewer: Michelle L. O'Donoghue, MD, MPH; Interviewee: Darren K. McGuire, MD, MHSc


January 10, 2018

Michelle L O'Donoghue, MD, MPH: Hi. I'm Michelle O'Donoghue. Joining me today is my friend and colleague, Darren McGuire, who is professor of medicine at University of Texas Southwestern and arguably one of the world's leading experts on diabetes. Welcome, Darren.

Darren K. McGuire, MD, MHSc: Thanks. It's great to be here, Michelle.

Dr O'Donoghue: It will be great to pick your brain on a few topics. Historically, we have not thought that diabetes meds were for cardiologists to prescribe, but I think more and more that is starting to shift. What are your thoughts about where we are going in this particular field?

SGLT2 Inhibitors and GLP1 Receptor Agonists in Cardiology

Dr McGuire: We are in a dynamic time. I've been interested in diabetes and cardiovascular (CV) disease for 20 years, and I've never managed blood glucose. It's not that I cannot, or that I want to shunt it off to other people, but it has not been in our domain as cardiologists because we have not had clinical outcomes data. Now we have a number of clinical trials of antidiabetes or antihyperglycemic therapies that have been proven superior for CV outcomes. It's pretty clear across these trials that [the outcomes] have nothing to do with blood glucose management.

We are kind of in a domain now, a bit like the statins, where diabetes has a certain amount of CV risk and therapies have been developed for the treatment of hyperglycemia that also have CV disease benefits that are independent of the glucose. We are going to have to start prescribing these. I already have.

With the sodium/glucose cotransporter 2 (SGLT2) inhibitors (ie, canagliflozin, empagliflozin, dapagliflozin), we have CV outcomes superiority for empagliflozin[1] and canagliflozin,[2] which have both been shown to reduce CV risk. Empagliflozin has a clinical indication in the product label to reduce CV mortality. These are once-daily tablets with nothing to monitor, and you do not have to titrate them. They are very well tolerated (there are a few side effects). These are pretty simple meds to prescribe.

Then we have this other class of injectable medications, the glucagon-like peptide 1 receptor agonists (GLP1 RAs). They are a little more challenging for the cardiologist. You have to do injection teaching, which we have begun to do in our clinic. We have [done injection teaching] for years for low-molecular-weight heparins, and now for proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors. Sometimes we teach patients how to inject themselves. For these GLP1 RAs, you tend to have to start with a low dose, and titrate slowly for gastrointestinal (GI) tolerability issues. It's a little more complicated.

I think it's time for cardiologists to start prescribing the tablets.

We are going to have to start prescribing these. I already have.

Patient Risk Versus A1c Reduction

Dr O'Donoghue: You raised some great points, because there might be some more reluctance on behalf of cardiologists to start prescribing the injectables simply owing to less familiarity. One of the interesting points you raised is that it's not clear that A1c reduction is correlating with what we are seeing with CV benefit. Correct me if I'm wrong, but there seems to be a divide: Endocrinologists are still largely relying upon A1c reduction when thinking about their algorithm for approaching different diabetes medications, whereas cardiologists are looking at the CV benefit and saying, "Hey, shouldn't we be thinking about SGLT2 inhibitors if not first line, then certainly near the top of the list of medications for our diabetes patients?"

Dr McGuire: That is exactly right. In the cardiology space, we are using the risk of the patient instead of glucose metrics to dictate or inform our clinical decisions. We have been doing that for years with statins—it's a very similar paradigm. It's not that glucose control is unimportant; for microvascular disease risk reduction, we have a number of therapies that have been proven safe but not incrementally beneficial for CV disease.

We can have confidence that some of these medications, such as the dipeptidyl peptidase 4 (DPP4) inhibitors (also once-daily tablets), have completely neutral effects on atherosclerotic CV outcomes.[3] They are very safe, with robustly proven safety in noninferiority trials. They do improve glucose control. Glucose control is a backdrop, and we can leave that to the endocrinologist.

Even the American Diabetes Association (ADA) has slightly disconnected the indication for the SGLT2 inhibitors empagliflozin and liraglutide, the once-daily injectable GLP1 RA. In the 2018 Standards of Care guidance from the ADA,[4] they recommend to consider both of those medications independent of blood glucose for patients with atherosclerotic vascular disease and type 2 diabetes.

Is CV Benefit a Class Effect of GLP1 RAs?

Dr O'Donoghue: You called out liraglutide specifically. Do you think CV benefit is a class effect for GLP1 RAs, or do you think it's restricted just to liraglutide?

Dr McGuire: It's a difficult question. We always have that long history of trying to deduce class effects across angiotensin-converting enzyme inhibitors and beta-blockers, and here we are with another class of medication with somewhat different signals from the four different outcomes trials. Pretty clearly, liraglutide improves CV outcomes.[5] In fact, it has a product label indication for reduction of CV risk that includes CV death, myocardial infarction, and stroke. It's the second medication in the history of diabetes now with a clinical outcomes indication.

Liraglutide and semaglutide[6] (which is given once-weekly) are very similar to native GLP1. The other medications in the class are quite different chemically. Extended-release exenatide[7] is based on the native GLP1 receptor agonist, but it's delivery system is a little bit different. Then we have the medication lixisenatide.[8] Those two medications have been shown to be safe, but not incrementally beneficial.

The point estimate for the large-scale outcomes trial with long-acting exenatide was in the favorable direction, barely missing statistical significance. That is a little bit suggestive of a class effect, but we have pretty clear data with liraglutide and semaglutide. I think those are probably going to rule the day, especially once we get to the once-weekly injectable. Patients really like that.

Choosing Between SGLT2 Inhibitors and GLP1 RAs

Dr O'Donoghue: Broadly speaking, if a cardiologist is watching this program today and is entertaining the idea of prescribing a medication from the SGLT2 inhibitor class or the GLP1 RA class, how do you approach which one you prescribe preferably over the other as first-line?

[I] never thought when I went into cardiology that I'd be having conversations with my patients about urinary hygiene.

Dr McGuire: My choice right now is empagliflozin. Canagliflozin data are impressive, but the problem is an incremental risk for amputation that comes with canagliflozin. Also, empagliflozin robustly reduced CV death. With the analysis of the canagliflozin data, that component of the composite outcome was not statistically different, and the magnitude of the difference for CV death was about one half of that seen with empagliflozin. If given a choice of modifying a three-point major adverse CV events outcome or death, I'm going to go with death.

Empagliflozin comes in two doses: 10 mg and 25 mg. All of the benefit from a CV perspective is with the 10-mg dose. I do not see any reason for a cardiologist to use anything other than 10 mg. We only have to learn one dose. That makes it simple. We do have to remind patients about urinary hygiene—I never thought when I went into cardiology that I'd be having conversations with my patients about urinary hygiene. The SGLT2 inhibitors do have a twofold to fourfold increased risk for mycotic genitourinary infections, yeast infections. You have to counsel the patients about that.

Finally, and importantly, we ask patients not to take these medications on any day that they are not taking a normal amount of oral intake, such as if they are sick, have GI distress, are not eating much, or if are fasting for a procedure or for religious purposes. This has to do with a very small but consistent across-the-class risk for diabetic ketoacidosis, which is a pretty rare complication in type 2 diabetes, but it can be seen with these medications.

Use for Primary Prevention

Dr O'Donoghue: Thanks for that useful information. We have data being presented on CANVAS[9] here at American Heart Association [Scientific Sessions], regarding the question of whether there is any difference in terms of CV benefit for prescribing SGLT2 inhibitors for the primary prevention or secondary prevention indication. Do you believe there is a difference?

If there is any benefit in the primary prevention realm, it's much more modest, and most of the time it's pretty neutral.

Dr McGuire: It's a great area of uncertainty. For example, of all adults with type 2 diabetes only about 20% have prevalent atherosclerotic vascular disease, and that is the patient population predominantly studied in all of these outcomes trials. There is a reason to believe that the incremental CV benefit may be restricted to those patients with some level of underlying disease and risk.

For example, in the EMPA-REG OUTCOME trial with empagliflozin, all of the patients had atherosclerotic vascular disease. In the two trials of CANVAS[10] with canagliflozin, about 70% had prevalent atherosclerotic vascular disease. In the trials that have enrolled both primary and secondary risk populations, it's those subsets with the prevalent atherosclerotic vascular disease that derive the benefit.

There have been subanalyses and analyses of statistical interactions, and those have typically been negative statistically. The reality is that across these trials, if there is any benefit in the primary prevention realm it's much more modest, and most of the time it's pretty neutral. That includes the GLP1 RAs semaglutide and liraglutide. It includes canagliflozin. It also includes the long-acting insulin, insulin degludec, that was compared against glargine.[11]

In all of these trials, the primary prevention subset really did not have a demonstrable benefit. These were small subsets—20%-30% of the patients—so the sample size was limited, and because it was a lower-risk population, there were fewer events to analyze. It just may be a matter of statistical power. It also may be a matter of the duration of observation for primary prevention. We may need to go 5 or even 10 years before we see these benefits.

What we know for certain is that it's patients with atherosclerotic vascular disease. Both the indication and the product label for empagliflozin and for liraglutide both are predicated on prevalent atherosclerotic vascular disease for CV risk reduction.

Dr O'Donoghue: That is very helpful. Certainly, patients with established coronary artery disease are going to have a higher risk for CV events. A lot of us have been grappling with the question of why there should be a differential benefit per se. Patients might have a lower event rate if they have not had a prior atherosclerotic event. You might still think there would be comparable benefit.

Dr McGuire: It's hard to understand how they may be differential. There are some hypotheses with the SGLT2 inhibitors that especially if you have underlying atherosclerotic vascular disease, you are more likely to have some kidney disease. It looks like through the mechanism of their inhibition of glucose reuptake from the urine, there are a lot of secondary effects on sodium handling, hematocrit, and circulating ketone bodies,[12] and other things that are beneficial or at least more beneficial in patients with atherosclerotic vascular disease. There is some reason to believe that may be a real differential treatment effect, but I would also turn it around and say that I'm very confident that these are safe medications to use in patients without atherosclerotic vascular disease.

If we go back to hanging onto the safety proof and understanding that long-term glucose control is beneficial, these drugs are really easy to take for most people. The SGLT2 inhibitors and the GLP1 RAs result in some weight loss, so patients love that. They have a little bit of blood pressure reduction and favorable effects on lipids, and they are safe. Even if we are not measurably affecting near-term CV risk, they are going to be very useful across the population.

Dr O'Donoghue: That is a great point to wrap up the discussion with. It will be very interesting to see over the next few years whether there is a shift in terms of the prescribing of diabetes medications toward cardiologists. The more we can inform people so that they are familiar with the data, the better. Thank you for helping to provide clarity on that point.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.