Freed: This is Steve Freed. We’re here at the American Diabetes 78th Association, Scientific Sessions, where all the people that are in the know when it comes to diabetes have to show up to get all the instant news and results of the studies. And today, we have a very special guest, John Buse, who has an illustrious career when it comes to the field of diabetes. Maybe we can start out with, tell us a little bit about your practice and who you are.
Buse: Yeah. I’m an endocrinologist by training. I did an MD PhD in immunology of diabetes and then did my fellowship research in Chicago with Dr. Graeme Bell on the genetics of type 2 diabetes. And pretty much after that I started doing clinical trials. So, the last 25 years I’ve been involved in these large multicenter clinical trials and pharmacoepidemiology type analyses around drug safety and efficacy.
Freed: So, before we get into the questions regarding why you’re here, with your background and scientific knowledge I always want to ask the question — and that’s one of the great things about what I get to do. I get to find out information before anybody else — what do you think is the greatest information coming out of the ADA at this session?
Buse: Well, I think of the sort of practical information, these new drugs, GLP-1 Receptor Agonists and SGLT2 Inhibitors are extremely exciting. They save lives. We have not had drugs that over short periods of time were associated with reductions in heart attack, strokes, and death, and they improve kidney outcomes. I think that’s the most exciting thing in diabetes in my career. I think the other thing that I try and pay attention to but I haven’t gone to any of the sessions yet is implementation. We have great drugs. We have great devices. We know what needs to be done but somehow as a society we don’t seem to seal the deal. And we need better programs at helping patients adhere to medications, helping patients do the right thing with regards to lifestyle intervention. I think the future is very rosy for folks with diabetes.
Freed: A cure?
Buse: A cure — well, for a type 2 diabetes, bariatric surgery is close for many patients. I think, well, I’m very impressed that these GLP-1 Receptor Agonists and SGLT-2 Inhibitors, mixed in with a little metformin and a little insulin, the vast majority of people with diabetes can have their diabetes controlled and live out normal life expectancies. Not a cure, but really good treatment.
Freed: So, let’s get into it. I think you had eight presentations.
Buse: I think so.
Freed: Let’s start with what is the OBSERVE 4D trial and why was it conducted?
Buse: Yeah. So, it’s actually not a trial. It’s a so-called observational study. It’s part of what we — which now is commonly referred to as real world evidence. The way the data is collected and analyzed is, it’s so-called, claims data. So, when a patient goes to a pharmacy and gets a particular drug, a code is recorded and that can be tracked through data sets; when patients go to the hospital for a heart attack, or for an amputation there’s a claim registered and coded, and that can be tracked in the database. And the pharmacoepidemiologist who led this study used four large administrative claims data sets involving millions of people and hundreds of thousands of patients with diabetes, and looked at the relationship of SGLT-2 drug use in individual patients and downstream outcomes with regards to heart failure and amputations. And the reason that’s of interest is the clinical trials have suggested that these SGLT-2 Inhibitors are really good for reducing heart failure hospitalizations. And the CANVAS Program with Canagliflozin suggested that there was a numerically small but statistically significant increase in the risk of amputations. So, in the study we have a positive control, the heart failure. And basically the bottom-line is the SGLT-2 Inhibitor use is not associated with amputation and the SGLT-2 Inhibitor use is associated with reductions and hospitalizations for heart failure.
Freed: How was the study designed?
Buse: It’s designed as a pharmacoepidemiology study. So, you get all the data and then you look for so-called new users, so these are people who haven’t been on an SGLT-2 Inhibitor before in the data set and then prescription for SGLT-2 Inhibitor drops in the data set for an individual patient, that person then becomes an SGLT-2 Inhibitor patient. And then the comparison was done basically against all other drugs in diabetes but using a similar design. And then they followed the people on SGLT-2 Inhibitors versus the people on non-SGLT-2 Inhibitors in the database. It’s not randomized like a clinical trial. It’s Dr. A prescribes to Patient B an SGLT-2 Inhibitor. Dr. X prescribes to Patient Y a Pioglitazone. And it’s just a mixture of all these people in the database, but because there’s so many numbers it’s really quite a powerful technique.
Freed: Amputation, is it a major concern? What were the findings for below-knee, lower extremity amputations and the overall study population?
Buse: So, in the overall study population there was no statistically significant difference with regards to the risk of amputations. The so-called point estimate of the hazard ratio was less than one, so the most precise estimate is that there is zero increase risk of amputations. The confidence intervals are wider, so we can’t rule out a small increase or a small decrease in the risk of amputations. And the thing that I like best about this study is it has this embedded control. So, hospitalizations for heart failure in the clinical trials, they’re reduced 40%-50%. Here in this data set, reduced about — the point estimate is reduced about 60%. So, the database works well for detecting an outcome associated with these drugs. And despite that, we weren’t able to demonstrate an issue with amputation.
Freed: And what did the subgroup analysis in patients with established cardiovascular disease show?
Buse: That’s a really important question. So, the CANVAS trial where the risk of amputation was first identified, that trial involved patients with prior cardiovascular disease. And so, we wanted to look specifically in the general US population reflected in these databases at the subset of patients who had clinical cardiovascular disease. And there was again no signal for a problem with amputations, and again evidence of a reduction in hospitalizations for heart failure.
Freed: Getting back to the amputation question, when the word came out that a particular drug in this class caused amputation, we automatically, at least the medical community assumes it’s for all the drugs in that particular classification. And the results are either a lot of physicians just hearing the word will not treat a patient with that particular drug. How do we overcome that misinformation that really affects the health of people with diabetes, and a lot of these prescriptions — just like Avandia, information came out, basically we stopped using it and now we find out that the information was wrong.
Buse: Yeah. And I wouldn’t say that the information is wrong. It just needs to be put into context. I think studies like this one where you look at millions of patients who were being treated as they are treated in the United States can provide some reassurance. And this isn’t perfect data, even though there’s large numbers of patient who are exposed to the drug. In the United States on average patients are only exposed for about six months, so many patients discontinue their drug because the insurance coverage switches or whatever. So, we don’t have a lot of patients, the data set had been exposed for two, three, four years like they were in the clinical trial. But I think physicians and patients can be reassured that at least in a real world setting there doesn’t seem to be a signal for amputations and there does seem to be a signal for those improved cardiovascular outcomes. Now, that said, I think if I had a patient who had had a prior amputation who’s at very high risk for having another one, let’s say they had lost a toe, they’re at very high risk for losing a foot or a leg, that patient would not be top of my list to use an SGLT-2 Inhibitor. And now, if they had heart failure, or something where we know the benefits are really high I might even steer them if we were going to use an SGLT-2 Inhibitor if they found this amputation question not so offputting, I might stir them to one of the agents that have not had clinical trial evidence for amputation like Empagliflozin or Dapagliflozin.
Freed: So, that brings up a good question. As a family practitioner where 90% of the type 2s go to, where should the physician go? Where do you go to get accurate information that you feel comfortable enough to take that information and put it into practice. If you put the word diabetes into Google, you get 32 million hits in about 1/10th of a second, so there’s just so much information out there right now and a lot of misinformation out there. Where should a physician go or where do you go to get accurate information?
Buse: Well, I go to the medical literature but I’m a diabetes specialist, I pretty much don’t read anything about anything unless it’s about diabetes, so I devote my life to it. It’s really complicated. And I go to my colleagues that are involved in these studies for help as well. It’s actually a great question of where the primary care provider should go to get that kind of accurate information and feel comfortable with it. The American Diabetes Association is a great resource, but their website, it’s not like they have a thing, “What would I do for a patient with neuropathy who’s had a prior amputation but has heart failure? And I think an SGLT-2 Inhibitor would be a really great drug.” There is no such resource. And frankly the best I could do or another diabetes expert could do would give you an opinion. Now, the problem is that we don’t have a lot of data with those kinds of details in them but this real world evidence over time will provide such data.
Freed: So, what were the results for hospitalization for heart failure in the overall study population?
Buse: Yeah. So, the heart failure hospitalizations with SGLT-2 Inhibitors have been well demonstrated and randomized, double-blind, placebo controlled clinical trials, so we know these drugs were for heart failure. In this study, in the database, the use of an SGLT-2 Inhibitor, Canagliflozin was the most commonly used one in the database, likewise was associated with about a, 60% is the best estimate, reduction heart failure. But basically the reason why we did it is just to see whether the database was adequate to find a signal for a benefit or a harm of the drug where we kind of knew already that there was this benefit.
Freed: That’s interesting. So, because there’s this benefit of reduction of hospitalization which is very expensive, do you find more insurance companies are opting to pay for these type of drugs to save money even though the drugs are kind of expensive?
Buse: It’s hard to generalize about insurance companies. There’s nothing you can generalize about insurance companies. And then within an insurance company they’ll have different plans. But I’d say in general these SGLT-2 Inhibitors are covered by most insurance plans. It may be at a higher co-pay than a generic drug, but the coverage is pretty good. And the results from the clinical trials, as I said when we first started talking for these SGLT-2 Inhibitors were absolutely stunning, reductions in heart attack, stroke, cardiovascular death. Things that people think are really important. And this additional benefit on hospitalization for heart failure and that’s clinical condition where there is no other drug that has shown this kind of benefit in a population of patients with diabetes and benefits on kidney outcomes. Again, of the complications that patients with diabetes face, it’s really cardiovascular disease and kidney outcomes that we don’t have a good handle on otherwise. The eye disease, we have laser therapy and all the miracles the ophthalmologist can do. For neuropathy, we have great efforts of podiatrists, and the foot care specialists to provide appropriate footwear. The heart failure, cardiovascular disease, kidney diseases, are really where the problems are for patients with diabetes.
Freed: The SGLT-2s, as we found out how it affects the heart and prevents death by 40% to 50%, I would think that we would be doing clinical studies for people without diabetes for SGLT-2s, is there such a thing?
Buse: There is. So, there are trials using SGLT-2 Inhibitors, big trials, specifically with kidney disease to sort of prove in a stronger way that these drugs can prevent the progression of kidney disease. There are also trials in patients without diabetes, but who have kidney disease to look at the same endpoints. And there are trials in patients with heart failure specifically designed to address, “Does this drug change the natural history of heart failure,” and that’s being done in patients with diabetes but also being done now in patients without diabetes. Mostly because the impact in the original study was so big and the problem is intractable otherwise.
Freed: What are the chances of putting an SGLT-2 in our water supply?
Buse: Almost. There are really — other than this concern about amputation, the other side effect is genital yeast infections. So, vaginal yeast infections in women, we call it balanitis. But infection of the foreskin in men, it’s a relative minority of patients. Most will have one episode and get over it. You could almost put it in the water supply. And I would just mention that there are people who are born with mutations in SGLT-2. They functionally auto SGLT-2 inhibited themselves. It’s a relatively uncommon mutation but they tend to not have problems with obesity, they tend not to have problems with diabetes, they tend not to have problems with kidney disease. They live out a normal life expectancy with less of those problems. So, maybe we should start talking about putting it in the water supply.
Freed: How did OBSERVE 4D trials or study compare to the results seen in other studies with Canagliflozin and other SGLT2s?
Buse: So, this isn’t the first such study. I mentioned before the clinical trials suggested that Canagliflozin might have a issue with amputation. Ertugliflozin is the other SGLT-2 Inhibitor that has amputation in the label of the drug. The other big clinical trial, the EMPA-REG study with Empagliflozin, didn’t demonstrate any problem with amputation. In this real world evidence space there are about three other studies that have been done before, but this is by far the biggest study that’s been conducted to date and it is specifically in the US population.
Freed: My favorite question to end this interview is since medical professionals will be watching this, what would you like them to take away from this, because they can’t be here, when it comes to all the presentations that you’re going to be doing here?
Buse: I think the most important thing on this topic is that the SGLT-2 Inhibitors have a big effect on cardiovascular outcomes and renal outcomes in patients with type 2 diabetes, at least those who have clinical cardiovascular disease. And this concern around amputations, at least in the way that the drug is used in the population in the United States, doesn’t seem to be a problem. Now that said, in patients who have had prior amputations or who have had neuropathy, high levels neuropathy and are at risk of amputations, again, it may not be a great idea to use these drugs upfront, unless there’s a compelling indication for benefit. Like in a patient who has heart failure or in a patient who has advanced cardiovascular disease, then you really have to weigh the potential risk and benefits. And that balance, even in those patients, seems to favor the use of these drugs to prevent heart failure and hospitalizations.
Freed: Although it hasn’t been approved, what about type 1s and the SGLT-2s?
Buse: So, there are a huge number of presentations here at the ADA on the use of SGLT-2 Inhibitors in type 1 diabetes and the main issue seems to be an increased risk of diabetic ketoacidosis. So, the SGLT-2 Inhibitors have efficacy in reducing hemoglobin A1C by about half a percent, body weight by up to 5% over a year, so a big benefit in weight, improvements in blood pressure, reductions in rates of hypoglycemia, at the same time you’re reducing average glucose, improved time and range, less glycemic variability, improved so-called patient reported outcomes. Patients just feel better on these drugs. So, all of that is very exciting but 3% or 4% per year additional risk of diabetic ketoacidosis. So, the regulators are going to have to think hard about this. I think in patients who are very careful in their self-care behaviors, very attentive to what’s going on in their body, I believe that in such a patient with type 1 diabetes, as long as they remember to check their ketones whenever they feel unwell, whether it’s a little bit of nausea or just like, “I don’t have any energy today.” Low level ketoacidosis is hard to sort out. But in general, those patients would have mild symptoms that would give them a hint, that they have ketones that really just needs to be treated with food, with carbohydrates, and insulin, So, it’s not like the treatment is so hard if it’s detected early. The problem is if you don’t catch diabetic ketoacidosis early people could end up in the intensive care unit.
Freed: Type 1s, because we’ve seen they do use SGLT-2s, so there’s a reduction in hypoglycemia and it’s due to the effect because my personal experience has been that they’re using a lot less insulin with the SGLT-2s, which means we can have less hypo just by virtually using less insulin.
Buse: Right. So, it is associated with the reduction in the insulin use. But by nature the SGLT-2 Inhibitors they work harder, they work better when the glucose level is high and they have less effect when the glucose level is low. So, by nature they tend to have sort of a buffering capacity where it does reduce glycemic variation.