DR. FONSECA: We'd like to move on to a slightly later stage of diabetes. And for that, I'd like to introduce Dr. Lawrence Blonde, who is the Director of the Ochsner Diabetes Research Unit in my home city of New Orleans, Louisiana. He has been very active in the area of diabetes clinical trials. He's an associate editor of Diabetes Care. He has served on the ADA board several years ago. He's very active and has been Chair of the Steering Committee of the National Diabetes Education Program, also co-chair and co-author for the recent AACE guidelines, and co-author of the recent algorithm that has come out of AACE. He's going to talk to us about GLP-1 agonists in patients with established diabetes. Welcome Larry.

LAWRENCE BLONDE, MD, FACP, FACE: Thank you. I'm going to talk about GLP-1 receptor agonists in people with a little bit more established diabetes. There will be a little bit of overlap with what Professor Nauck showed, but I'll try and amplify on some of the issues that he raised.

This is an audience response question. Which of the following is not true about hypoglycemia? Pseudohypoglycemia is hypoglycemia reported in error by faulty glucose meters. 2, DPP-4 inhibitors and GLP-1 receptor agonists are recommended when the goal of therapy is to avoid hypoglycemia. Or is it 3, 25% of patients with type 2 diabetes may experience nonsevere hypoglycemia daily to once a week? And 4, the risk of hypoglycemia increases with incretin-based therapies when they're used in combination with sulfonylureas.

Let's begin with a case. This is Phillip. He's 48 years old. He's an African American man. He's had type 2 diabetes for 12 months. He also has had hypertension. He has no history of cardiovascular disease, but he has had, in the past, increased LDL cholesterol. His medications include lisinopril, simvastatin, and metformin 1 gram twice a day. He is trying to follow the lifestyle intervention of increased physical activity and medical nutrition therapy.

His blood pressure is 138 over 84, so the diastolic is a bit above what is recommended by the ADA. The new guidelines from the ADA say the systolic should at least be less than 140. His LDL cholesterol is 98. His estimated GFR is 55. His height is 6-feet tall. He weighs 220 pounds, and he has a BMI of 30, which is considered obese. His A1c is 7.8%. His fasting plasma glucose is 140. His peak postprandial glucose is 180-220. His A1c is 7.8% on metformin.

What would you do next to improve his glycemic control? Would you, 1, add a sulfonylurea? 2, add a DPP-4 inhibitor? 3, add a GLP-1 receptor agonist? 4, add a TZD? 5, add a basal insulin analog? Or 6, add an SGLT-2 inhibitor?

Most people would add a GLP-1 receptor agonist but, appropriately, there's a smattering of other options here.

So, let's look at some of the data that can inform this. First of all, the question would be, "What would be the goal for this patient?" The American Diabetes Association says the goal for most people with diabetes should be less than 7, but as part of the personalized approach, they do point out the importance of individualizing care.

For some people, who have new-onset diabetes, have a long life expectancy, have a low risk for hypoglycemia, don't have established complications or many comorbidities, have a low risk for developing hypoglycemia or adverse consequences (from hypoglycemia), one could try and get closer to the nondiabetic range of about 6 percentage points. On the other hand, people with the reverse, people who have hypoglycemia or a risk for hypoglycemia or adverse consequences (from hypoglycemia), long duration of diabetes, short life expectancy, or established complications such as CVD, one has to accept higher levels of A1c as a goal. One of the first things you do is, along with the patient, establish what the goal would be.

We are doing better in terms of reaching some of these goals. These are data from the National Health and Nutrition Examination Survey conducted in this country over the years from 1988 to 2010. One can see that there is some increase in the percentage of people getting to an A1c less than 7, but still more than 40% of people are not getting there. You see improvement in patients getting their blood pressure under control. I think, surprisingly, you see more than 40% of people who have an LDL cholesterol that is above 100 and who are not on a statin.

If you look at the combination of the ABC's of diabetes, less than a quarter of people are getting there. There is evidence that we're doing better, at least with these studies. There's also some evidence from the CDC that complication rates are diminishing, but there is an opportunity for us to do even better and make an impact.

The cornerstone of therapy is lifestyle interventions. We all know how challenging it is to help our patients adhere to those. I always try and point out that the National Diabetes Education Program, a joint venture of the National Institutes of Health in the US and the Center for Disease Control and Prevention, have developed something called Diabetes Health Sense on their Web page. They got a group of behavior change experts to identify, about now, 160 resources that they think could be used by patients to help them better adhere to our recommendations, both for lifestyle interventions and for medications. If you haven't looked at this, I recommend that you do so. If you think it merits it, suggest it to your patients.

That being said, most patients are in fact going to ultimately require anti-hyperglycemic pharmacologic therapy in addition to lifestyle. For those who have no contraindication and who can tolerate it and who don't need insulin right at the time of diagnosis, the most frequently used agent is metformin. The ADA EASD position statement that Dr. Nauck talked about recommends this. Then if people do not get to goal on metformin, among the other choices, they would pick either a sulfonylurea, a thiazolidinedione [TZD], a DPP-4 inhibitor, a GLP-1 receptor agonist, or a basal insulin. Then we go on to triple therapy and to more complex insulin regimens.

As he pointed out, one of the key issues relates to hypoglycemia. This year, the American Diabetes Association, in collaboration with the Endocrine Society, put together a work group to address a number of key questions about hypoglycemia and diabetes. How hypoglycemia should be defined and reported. What are the long- and short-term outcomes? What is the impact on treatment targets? What strategies are known to reduce the risk for hypoglycemia?

They came up with some recommendations about the classification of hypoglycemia. They defined severe hypoglycemia as it has been generally defined before. For documented symptomatic hypoglycemia, they talked about typical symptoms with measured plasma glucose of 70 milligrams per deciliter or less. Asymptomatic, no symptoms but a measured plasma glucose of 70 milligrams per deciliter or less. Probable, typical symptoms but no measured plasma glucose. This relates to one of the questions: pseudo-hypoglycemia, which they said was an event during which a person with diabetes reports any of the typical symptoms but their measured plasma glucose is above 70 milligrams per deciliter, but approaching that level.

As pointed out, the recommendations where the goal is to avoid hypoglycemia include either TZDs, DPP-4 inhibitors, or GLP-1 receptor agonists added to metformin.

This is the algorithm that's recently been released by the American Association of Clinical Endocrinologists. They stratified the recommendations by the baseline A1c. They specifically recommended earlier combination therapy, and they had a larger group of agents, although categorized in a hierarchical fashion of preference, for adding to metformin, which they, of course, agreed was the initial choice for most patients as was recommended by the ADA EASD algorithm.

For agents that are associated with a low risk for hypoglycemia, in addition to TZDs, GLP-1 receptor agonists, and DPP-4 inhibitors, they would point out that SGLT-2 inhibitors, colesevelam, bromocriptine quick release, and alpha-glucosidase inhibitor are also associated with a low risk for hypoglycemia.

This addresses another issue about the questions that were asked initially. Here in a survey of over 400 US patients with nonsevere hypoglycemic events, they looked at, in both type 1 and type 2 individuals, the frequency of this occurrence. Obviously, in type 1 individuals it's very common, but in type 2 individuals, episodes of daily to once a week occurred in almost 25% of patients. When you answered the question, remember the question said 30%. So for the interest of the thing, why don't you consider 25 and 30% similar, as you evaluate that question. When they looked at people who had had at least 1 episode a month, it was almost 60% of patients.

It is more common than we sometimes have felt. These are data from a combined patient survey and retrospective claims database analysis. They looked at the impact of hypoglycemia, either no hypoglycemia, unconfirmed hypoglycemia, or confirmed hypoglycemia, on hypoglycemic-related cost, diabetes-related cost, and total healthcare cost. As hypoglycemia increased, there was an increase in both diabetes and total healthcare costs.

This, in another study, which is a prospective, case controlled study of almost 78,000 newly diagnosed patients with type 2 diabetes from the national Taiwanese database. They saw that, compared to individuals with no hypoglycemic claims, those who had 1 or more inpatient or outpatient hypoglycemia claims had a reduction in their long-term survival. There also was an increase, about a 2-fold increase, in the hazard ratio for cardiovascular diseases and all-cause hospitalization.

Moreover, so-called minor hypoglycemia, or even moderate hypoglycemia, has been shown in a number of studies to be associated with a negative impact on patient-reported quality of life, lower treatment satisfaction, more difficulty in adhering to therapy, and greater resource utilization.

This is a meta-analysis published last year on hypoglycemic risk with anti-hyperglycemic agents added to metformin. Compared to insulin, glinide, sulfonylurea, one sees that with DPP-4 inhibitors, GLP-1 receptor agonists, TZDs, alpha-glucosidase inhibitor inhibitors, there's really no increased risk compared to placebo. To this list one could add SGLT-2 inhibitors, colesevelam, and quick-release bromocriptine.

With DPP-4 inhibitors, when used as monotherapy in combination with metformin or pioglitazone, a very low incidence of hypoglycemia, which does increase somewhat, if you add it to sulfonylurea, other insulin secretagogues, or insulin, meaning, and as the package insert indicates, if you're going to add this medication to someone taking insulin or an insulin secretagogue, one should look at the A1c. If you're closer to target, one should consider reducing the dose or even eliminating the sulfonylurea or reducing the dose of insulin as necessary to reduce the risk for hypoglycemia.

Very similar data in these trials of hypoglycemia that occurred with GLP-1 receptor agonists with and without sulfonylureas in head-to-head trials. Only 2 cases of major hypoglycemia, and these occurred in patients who were also on a sulfonylurea. In 1 of these trials, in the LEAD-6 trial, there was less minor hypoglycemia with liraglutide compared to exenatide bid, even though liraglutide was associated with modestly greater reduction in A1c. Again, the same issue, if one is going to add a GLP-1 receptor agonist to the treatment of a patient taking sulfonylureas or insulin, one should consider reducing the dose of the sulfonylurea or insulin.

What about weight gain? As indicated, if you're trying to avoid weight gain when adding something to metformin, you're going to want to use GLP-1 receptor agonists or DPP-4 inhibitors, according to the ADA EASD algorithm.

In the AACE algorithm, they include other agents, including SGLT-2 inhibitors, colesevelam, bromocriptine quick release, and alpha-glucosidase inhibitors. In fairness, in the position statement from the ADA and EASD, they also include these in the text, but because they favored the use of the 5 agents that we talked about as an add-on to metformin, that's why, of those, when they talk about reducing the tendency to gain weight, they include only GLP-1 receptor agonists and DPP-4 inhibitors.

This is a meta-analysis that looks at weight changes with anti-hyperglycemic agents added to metformin. Again, you see with DPP-4 inhibitors, with alpha-glucosidase inhibitors, and GLP-1 receptor agonists, avoidance of weight gain. One would also see this with SGLT-2 inhibitors and the other agents that I have mentioned.

Let's look in a little more detail at what we would expect in our patient if we added agents to metformin. If you look at exenatide bid added to metformin, you see about a .9 percentage point, placebo-subtracted improvement in A1c.

With liraglutide in a study that didn't have a placebo, adding liraglutide to metformin was associated with a 1.5 percentage point improvement in A1c, which was greater than the improvement that was seen with sitagliptin. In another trial, adding liraglutide was associated with a 1 percentage point improvement in A1c, which was equivalent to the improvement seen adding the sulfonylurea glimepiride, but with the benefit of having less hypoglycemia and a weight benefit.

Meta-analyses of trials with GLP-1 receptor agonists show a mean of 2- to 2.4-kilograms weight loss with these agents. Generally, weight loss is similar with the different GLP-1 receptor agonists in head-to-head trials. There were 2 exceptions in a trial comparing liraglutide and exenatide once weekly. There was modestly better glycemic control and better weight loss with liraglutide. In a trial comparing liraglutide with the investigational agent, albiglutide, there was better weight loss and modestly better glycemic control with liraglutide.

I had the privilege of presenting a poster today on looking at the relationship between the weight response to exenatide once weekly and glycemic control. We divided the weight response into quartiles. The group that lost the most weight in that quartile lost 6 kilograms. The group that lost the least weight actually gained 1 kilogram. When we looked at the change in A1c, it was more similar across the quartiles of weight loss. It points out that, even for people who don't lose weight, there is still, for most people, going to be A1c improvement.

Remember, these agents are approved as anti-hyperglycemic agents. I think sometimes patients, and even clinicians, may have unreasonable expectations about the amount of weight loss. Even in people who don't lose weight there is likely to be substantial benefit from the A1c reduction.

This weight loss tends to be more fat mass than lean body mass.

Although some people have said, "of course you lose weight because the most frequent adverse reaction to these agents are GI side effects," in fact, in a meta-analysis of the liraglutide trials, even people who had no nausea had a significant amount of weight loss, a little bit more weight loss when they did have nausea.

Similarly, in a trial that involved both exenatide once weekly and exenatide bid, even people with no nausea lost a significant amount of weight. In a study that we reported a number of years ago, we provided evidence that it's unlikely that the majority of the weight loss is driven primarily by nausea.

In this slide, we're looking at the efficacy of adding DPP-4 inhibitor therapy to patients like our patient with type-2 diabetes inadequately controlled with metformin alone. You see about a range up to, for most of these trials, about .7 to .8 percentage improvement in A1c.

There are studies that have compared the use of a DPP-4 inhibitor and a GLP-1 receptor agonist. In this study, you see that, compared to sitagliptin in green, the patients who were taking either dose of liraglutide had greater reduction in both A1c and in weight. Similarly, if you look at a study that compared sitagliptin to exenatide once weekly, the patients taking exenatide once weekly had a greater reduction in both A1c and weight.

This is a study that was published this year, and I participated in this study. To be fair, this is a post hoc analysis of 2 different trials. The LEAD-6 trial comparing liraglutide and exenatide twice daily and the study that's already been shown comparing liraglutide and sitagliptin. In this study, we looked at those patients who were only on metformin and added either liraglutide, exenatide, or sitagliptin, and at the point of that addition, had an A1c of less than 8.

What you can see is that liraglutide had a greater reduction than did exenatide and had a greater reduction than did the DPP-4 inhibitor sitagliptin in that trial. Even though the starting A1c was about 7.6%. I think many people would have thought that one might have had a bit better response with the sitagliptin.

Related to this, you see more people getting to goal in these trials, the proportion achieving a composite endpoint of a glycosylated hemoglobin of less than 7 with no weight gain or reported hypoglycemia at week 26 was substantially greater with liraglutide than either of the 2 comparative agents.

Dr. Fonseca is going to talk more about this trial. I'm just pointing out that in patients, early use of incretins, particularly GLP-1 receptor agonists, may have benefit even in those people who ultimately end up needing insulin. In this study, patients on metformin not at goal, having the addition of liraglutide, 60% got to goal. They had a reduction in A1c and weight. The patients who didn't get to goal were randomized to either remain on liraglutide plus metformin or to add a basal insulin analog to that combination.

You can see that, when you did that, there was a further improvement in A1c. There was a reduction in FPG. Patients did not regain the weight that they had lost on liraglutide, although they didn't lose as much weight as the people who remained on metformin plus liraglutide. They had a relatively low incidence of hypoglycemia. It may be that by optimizing a patient's own endogenous glucose-dependent insulin secretion before you add exogenous insulin, one may have benefit.

I wanted to point out that, there are press releases from both the AACE in collaboration with the ADA and separately from the Endocrine Society, pointing out that the study by Singh dealing with pancreatitis was in fact a retrospective, administrative database study, which is not as robust as prospective trials. As pointed out, there were other risk factors that may have not been adequately accounted for.

The study didn't adjust for some disease characteristics, and the excess risk of hospitalization was small. Importantly, there are a number of other similarly constructed studies that did not come up with the same result. AACE, ADA, and The Endocrine Society just recommended, as is always appropriate, that people with diabetes should speak with their healthcare professional to assess which treatments are best for them.

Also, the prescribing information for all of the incretin-related agents, (pancreatitis has also been reported in patients taking DPP-4 inhibitors, as well as GLP-1 receptor agonists, and remember that a causal relationship has not been established), the labels make recommendations about those with a past history of pancreatitis. There's a little variation, but in general, for the GLP-1 receptor agonists, the labels say in people who have a past history of pancreatitis, we don't know whether taking an incretin agent would increase the risk. The label for the GLP-1 receptor agonists suggests that maybe one would do well to consider other therapies.

For patients otherwise who are starting on incretins, they should know about the signs and symptoms of pancreatitis. We should ask about their pancreatitis history. If they develop the signs and symptoms, they should discontinue these agents, along with any other potential contributory substances. If pancreatitis is confirmed, they shouldn't restart these agents, and they should report cases of pancreatitis to the FDA.

In summary, minimizing the risk and magnitude of hypoglycemia and/or weight gain should be a high priority when selecting anti-hyperglycemic agents for many patients with type 2 diabetes. GLP-1 receptor agonists and DPP-4 inhibitors have been highlighted as options for avoiding hypoglycemia and weight gain in recent algorithms. Using incretin-related agents, especially GLP-1 receptor agonists, to optimize endogenous glucose-dependent insulin secretion might help type 2 diabetes patients get to goal. If insulin is subsequently required, it might allow, although there are not really good data for this, a decreased insulin dose and, perhaps, a lower risk for hypoglycemia and weight gain.

Let's go back to our audience response question. Which of the following is not true? 1, pseudo-hypoglycemia is hypoglycemia reported in error by faulty glucose meters. Or 2, DPP-4 inhibitors and GLP-1 receptor agonists are recommended when the goal of therapy is to avoid hypoglycemia. Or 3, 30%, and for the purposes of this question you could use 25%, which is what the data really showed, of patients with type 2 diabetes may experience nonsevere hypoglycemia daily to once a week. Or 4, the risk of hypoglycemia increases with incretin-based therapies when they are used in combination with sulfonylureas. Which of these statements is not true?

We've actually had a significant change. The correct response is number 1, and almost twice as many people at the end got that correct.

So remember Phillip? On metformin, A1c is 7.8%. What would you add to Phillip's regimen? A sulfonylurea, a DPP-4 inhibitor, a GLP-1 receptor agonist, pioglitazone, a basal insulin, or a SGLT-2 inhibitor.

More people chose, and there isn't a single right answer here, more people chose a GLP-1 receptor agonist. With that, I'd like to thank you very much for your attention.