Hola from Barcelona, where the five-day-long European Society of Cardiology’s 2014 Congress – one of the largest medical meetings in the world with 30,000 attendees – just came to a close. There were some major news items from the meeting for the broader cardiovascular disease patient population, including a new Novartis drug that dramatically reduced adverse outcomes following heart failure that should be very promising for many patients with diabetes as well as a Sanofi/Regeneron PCSK9 inhibitor that shows promise for sizeable LDL reductions and possibly reductions in major CV events. Despite all that was going on in broader cardiology news during the latter half of the meeting, there was still plenty of learning focused more specifically on diabetes. Our top highlights from days #3-5 are included below.
1. In a Wednesday symposium on the cardiovascular safety of diabetes drugs, session chair Dr. John Cleland (Imperial College London, UK) expressed startling doubt about the connection between glucose lowering and microvascular complications. We were very taken aback by his remarks – probably more so than in any other conference in recent memory with the possible exception of Dr. Elizabeth Koller’s (CMS) comments at a DTS meeting last fall.
2. Dr. Giuseppe Rosano (IRCCS San Raffaele, Rome, Italy), member of the EMA’s Cardiovascular Working Party, shared his belief that we can be more cautious about approving new diabetes drugs because of solid existing therapies, and because the impact of glucose lowering on cardiovascular outcomes appears to be relatively modest. We certainly agree this is true in the short run; we point out longer trials are rarely funded so we have little to no chance to know whether this is really true overall.
3. Dr. Rosano suggested that the EMA’s decision to remove GSK’s Avandia (rosiglitazone) from the market may have had something to do with the way GSK shared (or, in his view, potentially failed to share) the full range of relevant clinical data on the drug’s safety.
4. Physicians, in Dr. Rosano’s mind, should take any potential future finding of cardioprotection from a diabetes drug CVOT with a grain of salt if the study was initially designed as a safety study.
5. In an interview with the esteemed Swedish diabeto-cardiologist Dr. Lars Rydén (Karolinska Institute, Stockholm, Sweden), we discussed the factors influencing key FDA decisions in diabetes as well as the safety of the TZD class.
6. In one particular session, reliance on A1c for diagnosing diabetes at the expense of using OGTT data came under heavy fire, with a consensus emerging that OGTT is the far more sensitive of the two tools.
7. Dr. Oliver Schnell (Helmholtz Diabetes Research Center, Munich, Germany) lamented that there are currently no diabetes drugs approved for the prevention of diabetes in Europe.
8. A rapid-fire abstract session featured data from Japan suggesting that pioglitazone might have beneficial cardiovascular effects following certain cardiologic procedures.
9. Some cardiologists are cautiously optimistic about SGLT-2 inhibitors’ chance to demonstrate cardioprotection, as we learned at a BI-sponsored satellite symposium on the subject.
10. Multiple ESC speakers agreed that heart failure does not receive enough attention in diabetes, and might deserve to be part of pre-specified endpoints in outcomes trials.
- Executive Highlights
- Top Ten Highlights
- Detailed Discussion and Commentary
- Symposium: Diabetes and Heart Failure – Liaisons Dangereuses
- Is Heart Failure Treatment Different in Patients With Diabetes?
- Is Heart Failure Treatment Different With Diabetes Patients?
Top Ten Highlights
1. Over the course of a symposium on the cardiovascular safety of diabetes drugs, co-chair Dr. John Cleland (Imperial College London, UK) shared some startling personal perspectives on diabetes. He called into question the diabetology mantra that there is a strong link between better glucose control and reductions in microvascular complications, suggesting that the data supporting that theory is “based on one trial on type 1 diabetes patients with insulin, along with a few scraps of information elsewhere,” evidence which he does believe is strong. He went on to suggest that HCPs should perhaps stop treating diabetes until fasting glucose levels rise to the UKPDS level of 15 mmol/l (~270 mg/dl!), as that would save money through reducing the size of the diagnosed population; this would be a temporary move until better trials could confirm the connection between glucose lowering beyond and reductions in microvascular disease (this of course would itself take years to do, although he did not acknowledge this). Curiously, Dr. Cleland also seemed to believe that obesity and diabetes cannot be overlapping conditions, and suggested removing obese diabetic patients from studies on diabetes “because they’re not really diabetic – they’re just fat.” Dr. Cleland may have been selected as a co-chair with the hope that he could provide fresh perspective as a non-diabetologist, but ultimately we found some of his arguments to be quite counterproductive to the discussion in this otherwise quite illuminating session.
2. Earlier in the session, EMA Cardiovascular Working Party member Dr. Giuseppe Rosano (IRCCS San Raffaele, Rome, Italy) shared his fairly cautious personal perspective on the approval process for diabetes drugs. In a fairly Nissen-esque message, Dr. Rosano suggested that the medical community has been misled by an overly glucose-centric mentality. Compared to a few decades ago, when the incidence of diabetes complications was generally much higher, Dr. Rosano believes that today we can afford to be more cautious when approving new drugs for diabetes. Increased selectivity is also an option because of the choices we have available, including metformin and sulfonylureas – Dr. Rosano suggested that slow-release formulations of SFUs like glimepiride are unlikely to cause much hypoglycemia, and that positive hypoglycemia claims vs. SFUs for newer drugs have been based on comparisons against the standard release formulations (presumably weight gain and short duration are still an issue).
3. Later in his talk, when discussing the events that precipitated the FDA and EMA’s newfound scrutiny of cardiovascular safety evaluation for diabetes drugs, Dr. Rosano suggested that the EMA forced the withdrawal of GSK’s Avandia (rosiglitazone) not solely because of the safety data, but also because of the way GSK appeared to be hiding that data. More broadly, Dr. Rosano wondered aloud whether drugs like Avandia should have been widely used in the first place given that they exhibited significant increases in heart failure risk, an endpoint that Dr. Rosano believes deserves a greater place in the FDA’s cardiovascular safety evaluation process. There are many complexities here of course, not the least is that heart failure is defined so differently by so many.
4. Dr. Rosano emphasized that outcomes studies that are initially designed to confirm safety should not be seen as reliable sources of information on cardioprotection. Dr. Rosano forecast that in the next three to four years, the scientific and medical community will potentially be bombarded by results from outcomes trials, some of which could show a cardiovascular benefit. The fact that CVOTs are generally designed as safety trials rather than superiority trials, in Dr. Rosano’s view, should color our perception of the results, lest we be swayed by messages that come more from companies’ marketing departments than from trial investigators. We found some of the implications on the whole insulting though we agree that such trials may not be so valuable in determining whether compounds are cardioprotective; in our view, this is true because the trial length likely won’t be long enough, in most cases, to establish cardioprotection, even though it may be shown after a longer period of time. We agree that safety trials aren’t the best way to design trials that could show cardioprotection.
5. We had the chance to interview the esteemed Swedish cardio-diabetologist Dr. Lars Rydén (Karolinska Institute, Stockholm, Sweden), covering topics such as the drug regulatory process in the US and the data on TZD’s safety. Dr. Rydén lamented that science is slowly being pushed out of the driver’s seat by legal and political concerns when it comes to decisions to continue or discontinue major clinical trials (such as the TIDE trial on Avandia [rosiglitazone] and Actos [pioglitazone]). Dr. Rydén firmly believes that more cardiovascular outcomes data for diabetes drugs is needed, an idea which he believes must be applied to old diabetes drugs as new drugs. He argued that if society demands that pharmaceutical companies invest in CVOTs, then society should be willing to bear the cost of new outcomes trials for old agents such as sulfonylureas – we heartily agree with this. See below for a full transcript of the valuable interview / discussion.
6. In a symposium on diabetes and the EUROASPIRE IV study, over-reliance of A1c for diagnosing diabetes came under fire in a big way. The consensus was that two-hour oral glucose tolerance tests (OGTT) are a more sensitive way to identify diabetes patients. Dr. Jaakko Tuomilehto (University of Helsinki, Helsinki, Finland) concluded his presentation by suggesting that taking A1c alone can be “basically a useless expense” if done alone. We felt this was overstated. Subsequent speaker Dr. Oliver Schnell (Helmholtz Diabetes Research Center, Munich, Germany) connected the inadequacy of A1c alone to a more familiar (of late) topic: the importance of glycemic variability. He noted that short-term hyperglycemia can be harmful to the endothelium, and that the early perturbations in postprandial glucose that might be the earliest signal of type 2 diabetes cannot be picked up by A1c. In an interview we included in our ESC 2014 Days #1-2 Report, Dr. Naveed Sattar (University of Glasgow, Glasgow, UK) discussed the other side of the issue, noting that A1c (despite its limitations) is quicker and more feasible for a wider range of providers to do, adding to its value as a diagnostic for diabetes. Simplicity is very key in our view when addressing large numbers of people.
7. Dr. Schnell fielded an instructive question on pharmacological diabetes prevention, lamenting that there are no drugs approved as of yet in Europe for a diabetes prevention indication. He noted that both metformin and alpha-glucosidase have shown promise in preventing the progression from prediabetes to diabetes, and that as a result they have been approved in some Asian countries for prevention indications. Lifestyle remains the best diabetes prevention tool available to patients in Europe due to the resistance of local reimbursement and regulatory authorities such as the EMA in Europe, NICE in the UK, and IQWiG in Germany to approve drugs for prediabetes. Dr. Schnell believes that drug approvals for diabetes prevention would be valuable, as many patients cannot achieve the results they desire with lifestyle alone.
8. The conference’s primary diabetes-related oral presentation session, “Diabetes Under Siege,” featured some valuable learning in the form of five-minute rapid-fire talks. In a Japanese study, diabetes patients taking pioglitazone who underwent implantation of a drug-eluting stent were found to have a statistically significant ~50% lower risk of major adverse CV outcomes than diabetes patients not on pioglitazone that had the same procedure. Although this trial studied a very specific set of patients, the results could be seen as aligning with the more broadly applicable results of the PROactive trial, which found a 16% trend towards a reduction in MACE with pioglitazone. Presented in the same session, a sub-study of the population-based Rotterdam Study found that patients that engaged in mind-body practices such as meditation and yoga were 32% less likely to be diagnosed with the metabolic syndrome.
9. As we saw during a satellite symposium on the potential for cardioprotection with SGLT-2 inhibitors (sponsored by BI and the EASD's Diabetes & CVD Working Group), some cardiologists are cautiously optimistic about this newest diabetes drug class. This sentiment is contrary to the more cautious-to-pessimistic characterization of diabetes therapy we generally heard at the meeting. We can see why cardiologists might be slightly more positive towards SGLT-2 inhibitors, as the class has impacts beyond glucose on a number of parameters that have more robust impacts on cardiovascular risk such as blood pressure and body weight.
10. A Monday morning session hammered home the point that heart failure is not receiving enough attention in diabetes. Epidemiological data presented by Dr. Hector Ventura (Tulane University, New Orleans, LA) showed that diabetes patients are at greater risk for heart failure, with the risk rising approximately 15% per 1% increase in A1c. Multiple speakers at ESC (especially those that specialize in heart failure, unsurprisingly) called for heart failure to be a major adjudicated and pre-specified endpoint in clinical trials for diabetes drugs. At diabetes conferences, we have heard endocrinologists (such as Dr. Philip Home at CODHy Latin America earlier this year) suggest that heart failure is a difficult endpoint to adjudicate consistently across geographies. Although we found ESC speakers’ argument that heart failure is important for diabetes patients to be fairly compelling, it is also the case that CVOTs for diabetes drugs might suffer if additional complicating factors and burdens are placed on them – this is particularly true since as we understand it definitions for heart failure vary fairly widely.
Honorable Mention: Partial results from ADVANCE-ON, the long-term follow-up of patients from the ADVANCE outcomes study, were presented at ESC – although the results focused on the impact of blood pressure, seeing this data whetted our appetite for the diabetes results, which are to be presented at EASD. The impact of blood pressure reduction during the ADVANCE trial on cardiovascular outcomes and mortality was attenuated somewhat during the follow-up. However, blood pressure appears to have a more responsive impact on cardiovascular outcomes than glucose lowering (at least based on recent outcomes trials), and so we remain hopeful that the data presented at EASD will show evidence of a legacy effect from intensive glucose lowering.
Detailed Discussion and Commentary
Lars Rydén, MD, PhD (Karolinska Institute, Stockholm, Sweden)
During the third day of the conference, we got the opportunity to interview leading European diabetes-focused cardiologist Dr. Lars Rydén, whose multiple podium appearances at this year’s ESC included a presentation on the potential for cardioprotection with SGLT-2 inhibitors as well as a talk on lessons learned from the TZD class. Our interview touched upon these two topics as well as a number of others – on TZDs, Dr. Rydén lamented that science is slowly being pushed out of the driver’s seat by legal and political concerns when it comes to decisions to continue or discontinue major clinical trials (such as the TIDE trial on Avandia [rosiglitazone] and Actos [pioglitazone]). Dr. Rydén firmly believes that more cardiovascular outcomes data for diabetes drugs is needed, an idea which he believes must be applied to old as new drugs. He argued that if society demands that pharmaceutical companies invest in CVOTs, then society should be willing to bear the cost of new outcomes trials for old agents such as sulfonylureas, which could very well show cardiovascular harm in a long-term prospective study. On the topic of CVOT design, Dr. Rydén wisely pointed out that interventions will have the best chance of showing cardioprotection when they are tested in patients without established cardiovascular disease, although he acknowledged the difficulty in finding a balance between keeping studies feasible and studying patients in which prevention is a possibility. Read on below for the full transcript of the interview, and stay tuned for our coverage of more ESC sessions that benefitted from Dr. Rydén’s presence.
Interview with Dr. Lars Rydén: Full Transcript
Q: Hello Dr. Rydén, it is so nice to be able to sit down with you today. You have quite a portfolio of recent work and publications – we would love to hear what your current project is.
A: We are just now working on the EUROASPIRE IV survey, which is a survey of many thousands of patients with established coronary artery disease in 24 European countries. We looked at data from six months to a few years after the event in an attempt to find out how well they were treated for and informed about their disease. Linked to that project was also a careful evaluation of whether they have diabetes. All of the enrolled patients are investigated in terms of HbA1c and fasting glucose. If they do not have known diabetes, we would also do an oral glucose tolerance test. The first part of this survey that was in in-hospital patients or patients with established coronary artery disease. That has been completed, and the report has been written.
Now, we are working on a similar survey in primary care, where we are looking at patients who may be considered at a higher-than-normal risk of cardiovascular disease because of, for instance, hypertension, obesity, or high blood lipids. We also determined whether these patients have impending or established diabetes and perform oral glucose tests so that we can discover previously undetected glucose perturbations. This will also comprise five to eight thousand people from 19 countries in primary care across Europe to get as representative sample as possible. In the first part of the study, the hospital-based part, we have actually found an improvement in the utilization of various drugs, such as aspirin, statins, and blood pressure lowering drugs. However, use of the combination is still not at the level it probably should be. If you look at patients with and without diabetes, particular those with diabetes, you find that patients are not as well cared for as they should be. Many of them still have much higher blood pressure than what is recommended in the available guidelines. Blood lipids are not controlled to the extent they should be. Some patients are perfectly controlled, but quite a few have rather high levels. Glucose control is not perfect. Most patients were on one or even a couple drugs, but relatively few were on the entire combination, which means that there is room for further improvement in treatment.
If you translate that into recent data from, for example, the Swedish Myocardial Infarction Registry, which has followed all patients who have had myocardial infarctions under the age of 18 in Sweden for many years, we can see a successive decline in one-year mortality and that is valid both in people with and without diabetes. But over the years, from 1994 to the most recent data in 2012, there is about a 30% higher mortality in people with diabetes. Part of that gap probably relates to poorer treatment than it should be, so there is still room for improvement. And the interesting thing is you can probably still make improvements with available drugs. You don’t have to search for new ones. You can use the ones you have, which are not very expensive, but use them better. And you have to consider how patients are cared for. Some are looked at by general practitioners; some by cardiologists; some endocrinologists; some internal medicine specialists. But it seems that each patient may be seeing one of two different physicians, and one of the reasons they are not extremely well cared for may be that each physician thinks that the other one should take care of a given problem. Relatively few patients are at specialist nurse-based care centers or at rehabilitation centers where you have a holistic view that can take care of everything. Perhaps, better future care will involve a general practitioner who takes care of the patients, takes complete responsibility, and if there are problems with blood pressure or glucose or anything else, they refer patients to specialists to get support.
Another thing my research group is doing is that we are looking at the Swedish Registry for heart failure patients, splitting them between those with and without diabetes, and looking at the reasons for the heart failure and how well they are treated. There are a couple posters here at this conference where we show that diabetes-related cardiomyopathy doesn’t seem to be a very big issue. Ischemic heart disease and hypertension are the prevailing reasons for heart failure in people with diabetes and, despite that, relatively few of them have been investigated with coronary arteriography. We think that there is a gap here. Perhaps people think that people with diabetes are vulnerable and difficult to take care of, so we shouldn’t do too much or undertake treatments that are too complex. I think that is wrong. I agree that these patients are vulnerable, they do have a poorer prognosis. However, that also means that if you can treat them, the number needed to treat to save a life or prevent a myocardial infarction is relatively low.
Q: Given the news stories that have come out over the past few years about the TZDs, patients and providers lack certainty about the class’ safety. How do you think about the safety profile of the TZDs?
A: I’ll present a lecture on what we have learned from TZD trials on Wednesday morning. In principle, there’s only one TZD available now: pioglitazone. Pioglitazone has been suspected for an elevated risk of bladder cancer, and that is not taken out of the blue, because in animal experiments, there are actually indications of bladder tumors. There is also a French registry study – and some others – indicating that the number of people who develop bladder cancer on pioglitazone is higher. However, those are registry studies and for what they are worth, they can only generate hypotheses. So we don’t know for sure, and that caused the FDA and even the EMA to raise a warning that you should be cautious with patients who for other reasons have an increased risk of bladder cancer. Rosiglitazone is not on the market any more and aleglitazar is a balanced PPAR-alpha-gamma agonist. The discontinued AleCardio study had a negative outcome – negative in the sense that there was no impact on cardiovascular events in people with acute coronary syndrome and established or newly detected diabetes, and there were also a number of side effects, which were not unexpected because we know that glitazones cause water retention that can provoke heart failure. We know that may cause fractures, especially in women. We know that kidney function deteriorates to a certain extent, but we know it comes back again; if you just stop the drug, the kidney function normalizes. All of this caused Roche to abandon this project and just quit everything with cardiometabolic impact. And now, there are no new TZDs in the pipeline, and I don’t think there ever will be. The problem with the TZDs the first time was actually addressed by Steve Nissen and related to rosiglitazone. In the TIDE trial, which I was on the steering committee for, we wanted to make a head-to-head comparison of pioglitazone, which had some positive effects in the PROactive study in terms of cardiovascular events, and rosiglitazone, to see whether there was a difference. However, the FDA stopped the study about two years ago. I think that was a bad mistake. And one year thereafter, they said, “Well perhaps we over-exaggerated the risk with rosiglitazone,” but now it was too late because it was already taken off the market in Europe and partly in the US. So a very interesting class of drugs that has more or less been abandoned.
And why is this sad? Because glitazones, in particular the balanced alpha-gamma ones, impact a number of things that we know are dysregulated in diabetes. They control glycemia, they increase insulin sensitivity, they have a certain reduce the effect of high blood pressure, they increase HDL cholesterol, and they have an anti-inflammatory capacity. All these effects have been demonstrated in phase 2 clinical trials in, for instance, aleglitazar. However, when you test it in the drugs in the real world with hard cardiovascular endpoints, it still failed although we get all these beneficial effects. So what is the solution for people with diabetes? Obviously, it’s not that simple. We know that lipid disturbances, hyperglycemia, pro-inflammatory, pro-coagulant effects in these patients actually comprise a risk. But a drug that influenced all of that really would not help them. We need completely new drugs. We need a statin for diabetes, like we have a statin for lipids.
Q: Regarding aleglitazar, was it a case where the compound was doomed because of it’s clinical profile or was it a way the trials were designed in phase 3?
A: No. I think it is actually a case of a drug that seemed to be very effective, but studies conducted in the real world were negative or neutral. Because it was stopped prematurely, if there were any beneficial effects in subgroups, we cannot really answer that question. However, the drug did exactly what it was expected to do, and it didn’t help. You can ask yourself: Is the population you have addressed, people with acute coronary syndrome and known or newly detected diabetes, an ideal population? Perhaps not. Other trials and meta-analysis have generated hypotheses that if you start late, when the cardiovascular impact or injury is already established, it is too late. So perhaps the study population with established coronary artery disease was not an ideal population. Researchers should consider trying to detect prediabetes at an early stage and prevent the atherosclerotic complications from ever developing. In China, there was a study, called the Da Qing study, in which researchers randomized people with prediabetes (i.e., impaired glucose tolerance) to lifestyle intervention. This intervention consisted of a composite of increased physical activity and reduced dietary intake or calories. After six years, they showed that the development of diabetes in the intervention group with prediabetes was retarded compared to those in the control group where people didn’t do anything. After 20 years of follow up – this was published this year actually – they showed that the incidence of severe retinopathy is significantly less in the intervention group. They’ve also shown that cardiovascular mortality – and due to that, even total mortality – is reduced in the intervention group. So a moderate increase in physical activity – at least three hours a week of brisk walking or something similar – and a moderate weight decrease – let’s say five to ten percent – not only delayed the development of diabetes, but also reduced the development of macrovascular and microvascular complications. That study suggests that if you start aleglitazar in a population that doesn’t have established cardiovascular disease, then perhaps the drug would have helped. We don’t know because it’s never been done. However, in the future, trials attempting to stop diabetes complications in the macrovascular or microvascular world have to start very, very early, and the treatment has to be a composite of lifestyle intervention and a drug.
Q: In terms of the next “statin for diabetes,” where should we be looking? Where could something extremely transformational come from?
A: If I could give you a straightforward answer to that, I would be up for the next Nobel Prize in Physiology or Medicine. I don’t know. Obviously, it is not enough to interact with risk factors that are obvious. Perhaps the risk factors are actually not the pathophysiological key to the problem. Perhaps they are an expression of something. Perhaps insulin resistance should be counteracted. But that was done in the TZD trials because they improved insulin resistance, though it is possible that the intervention was too late. The newest compound, which is coming out as maybe a roaring lion, or at least a little kitten, is the SGLT-2 inhibitor class. The average weight reduction in the trials is about two or three kilograms, long lasting. At the same time, you have osmotic dieresis and other factors that together with the weight increase reduce the blood pressure. So, you have two things that are positive. The side effects so far are few, if any; maybe urinary tract infections, but they can be handled. There are a number of animal experimentation observations that are quite interesting, but not studied to the extent they need to be. For instance, a very recent study showed that at least half of the weight reduction is visceral fat. Visceral fat is responsible for producing leptin, which increases the problems with insulin resistance and diabetes complications.
So, there is a completely new class of drugs that are coming in through the door. It’s completely new. It may be in the future that it can help us. But we had the same expectations with TZDs. What must be done now is that we must use this novel class of drugs in suitable populations as rather than in populations that are desperately ill. Why do researchers choose patients with established coronary artery disease and myocardial infarction, etc.? It’s because they want to have a high number of events. After all, if you don’t have a high number of events, you have to have enormous populations or very long periods of observation. If you take a patient at an early start, it takes some time to before the events appear. So it’s a tricky business to do these studies. But, here with this new class we have the opportunity.
Q: Who should we be looking to in terms of who will fund these longer outcomes studies in patients with less advanced complications for whom prevention is more realistic?
A: We are talking about a global problem of enormous magnitude. The combination of diabetes and cardiovascular disease will cause a lot of suffering and a lot of costs. I think that drug companies have to earn money and they have now been frustrated over the effects of the TZD trials as well as the FDA regulations. I think that for instance, we do have compounds like sulfonylureas that have never ever been tested in the same way as the new drugs. SFUs, I think, have theoretically or pathophysiologically have some rather bad effects in increased weight. If the heart gets ischemic, it may very well stop the heart from protecting itself from ischemic injury. There are registry studies from Denmark and elsewhere showing that people on SFUs have a higher morbidity and mortality in the long run, but nobody questions SFUs. They are a cheap drug and it’s commonly used. I think if the FDA demands very accurate outcomes studies for SGLT-2 inhibitors, TZDs, or any new drug, it should also demand them even for the SFUs. But who is going to pay for that? Of course, the society because if the society puts a demand on pharmaceutical companies in some respect, they should be able to put out the same demands on themselves, but they don’t and this is very bad. I want to see a head-to-head comparison between these types of new and old compounds, for instance. So it’s mixed responsibility, I would say.
But meanwhile, we can go back to where we started. Before the first European guidelines in 2007 came out, I did a study together with collaborators called the Euro Heart Survey on diabetes and the heart, showing that many people with diabetes and cardiovascular disease didn’t know about their diabetes. This showed that for people who are on truly good evidence-based treatment and who get coronary intervention immediately when needed, their prognosis improved considerably. In that study, we were able to show – but it’s an observational study so we couldn’t control for everything – but the prognosis for people with diabetes improved and became very close to people without diabetes, but only if they got the full-blown treatment. Now we have shown recently that in the new Euro Heart Survey, there has been improvement of treatment but that there is still a gap between what the guidelines ask you to do for your patients and what physicians – around Europe at least, but perhaps also in the US – are actually doing. So while we are waiting for the new wonder drug, perhaps we could start using ordinary treatment – which is cheap, available, usually safe – in the proper way.
Q: Moving on specifically to the design of CVOTs for diabetes drugs, one of the challenges that is becoming more recognized is that of interim data disclosure, which is sometimes used to support submissions in the US. It seems like this is the agency’s balance between trying to getting data on safety but also not slowing down the process of approval. Is that a necessary compromise or are there better ways to get a look into a trial rather than unblinding a trial in the middle?
A: In principle, I think if you design the trial on some scientific grounds, with good pre-trial studies and with a solid hypothesis, then the trial should usually go on until it has reached an end. This could either be that the Data Safety Monitoring Board says that it’s no use to continue or like in the heart failure trial here, where we have a 20% mortality reduction earlier than expected, then you should stop the trial and give patients the drug. But it could also be that you have an over-mortality or over-morbidity in which the drug will not help, but rather cause problems. Then you should stop the trial. That is the DSMB’s decision. They are expert scientists. But like in the TIDE trial, when it was asked if rosiglitazone is really truly dangerous and if pioglitazone is safe, we started the trial and recruited thousands of patients and went on with our trial. And then by some legal, security, or economic reasons, the FDA was cowardly and they voted against the experts’ opinion. the experts said that the trial needed to continue, but the FDA stopped it because they were afraid of legal implications. Then it is not science anymore. There are many other arguments, which are not based on true knowledge and true sound scientific grounds. So that’s a very bad example. Things shouldn’t be like that. The aleglitazar study was stopped prematurely due to lack of efficacy and the company decided to abandon the project. I personally didn’t like that decision because we didn’t hurt people and although we didn’t really save lives, we had some indications that in some subgroups, it could be worthwhile. And we knew that some subgroups were more vulnerable. So let’s say that we continued the trial to the end. We could say that TZDs should be used in this type of patient and not in this other type. We cannot answer that question anymore because the study was prematurely stopped. So it’s an enormous investment and then comes out not being useful in the sense that it should have been. Presently, I think that academic people and scientists are not in the driver’s seat anymore when it comes to things like that. It’s economics and legal aspects that have the two large impacts. They should have an impact and safety is a very important issue but the safety should be in the hands of those who know about trials and not those who want to cover their backs for possible legal accusations. The insurance of US physicians is about my salary. And of course, they get a high salary to afford that, which makes healthcare costs awful, unnecessarily so.
Q: On that issue or more broadly, what are some of the things regarding diabetes care, diabetes research, or regulatory factors involving diabetes that you think the US could learn from Europe?
A: Well, the US has a complex situation because of all the lawyers and politicians. Lawyers earn a lot of money asking the patients if they have anything to complain about. They say, “let’s sue and we will share the surplus if we get something and I will be free of cost for you, etc.” That is an awkward system. We tested a system like that in Sweden and it was abandoned, forbidden. I think that the FDA regulations are fine. We need such regulations. But when the FDA asks for something, they have to stick to that. When science is not gambling, some companies will have a compound and are actually willing to start a phase 3 clinical trial without proper background research. That didn’t happen when I was young. We did lots of small trials, lots of building up knowledge, and then went out to the clinical trials. Now everyone wants to be first with their new wonder drug, which means they jumped over a number of steps that are necessary. Then they go to the FDA and ask, “if we do this, will you approve our drug?” I think that is not really as it should be. But the difference between the US and Europe in that respect is not large. The EMA in Europe is the equivalent of the FDA and they are much of the same thinking. And they are needed, but they need to listen to the scientists more and to the lawyers less.
Symposium: Diabetes and Heart Failure – Liaisons Dangereuses
Hector Ventura, MD (Tulane University, New Orleans, LA)
Dr. Hector Ventura began this symposium with some epidemiological data to underscore the importance of heart failure for diabetes patients. The UKPDS reported heart failure incidence rates of 2.3 to 11.9 per 1000 patient-years during the 10-year follow-up, and every 1% elevation in A1c raises diabetes patients’ risk of heart failure by approximately 15%. Predictors of heart failure in diabetes patients include higher A1c, elevated blood pressure, and poor renal function. Overall, Dr. Ventura concluded that diabetes patients are at clear elevated risk for heart failure, and that diabetes patients that experience heart failure are more likely to have a worse prognosis in terms of the length of their hospitalization and other parameters.
Questions and Answers
Q: In one of your slides, you showed that “insulin-dependent” diabetes had an adverse prognosis. When using that term, did you mean type 2 diabetes treated with insulin, or type 1 diabetes?
A: That was data from ADVANCE, so it was type 2 diabetes patients on insulin. Patients on insulin tend to have had a longer duration of diabetes, which could have been a factor explaining that effect.
Dr. Lars Rydén (Karolinska Institute, Stockholm, Sweden): The most we know about heart failure and diabetes is in type 2 diabetes. Type 1 diabetes is much less studied, and should receive more study. It is also important to remember that patients with prediabetes are at increased risk for many endpoints.
Rudolf de Boer, MD (University of Groningen, Groningen, Netherlands)
Following Dr. Ventura’s presentation that established that heart failure incidence is elevated in diabetes patients, Dr. Rudolf de Boer discussed possible pathophysiological mechanisms to explain the epidemiological connection. He divided the potential causes into three main categories: associated comorbidities, accelerated atherosclerosis, and the somewhat “mystic” entity of diabetic cardiomyopathy. Dr. de Boer indicated that a very large proportion of the excess heart failure risk in diabetes is due to the comorbidities of diabetes – a presentation earlier at the meeting discussed a similar topic of how macrovascular disease frequently occurs on a background of microvascular disease and its complications. Regarding diabetic cardiomyopathy (a term for diabetes-related increases in heart disease in the absence of underlying coronary artery disease), Dr. de Boer said that although a wide body of literature exists, the exact pathophysiology remains to be elucidated. Compared to the association between diabetes and heart failure, the connection between glucose-lowering therapies and improvements in heart failure is much less established, with only limited early-stage evidence for certain drugs (e.g. metformin and the DPP-4 inhibitor vildagliptin) that may reduce cardiac fibrosis.
Questions and Answers
Q: You expressed some doubt about the evidence of diabetic cardiomyopathy. Do you believe that diabetic cardiomyopathy is just a stage of development of heart failure, or might it be a separate entity?
A: I think there is a discrepancy between how it is being portrayed as a separate entity, yet there is a lot of discussion on what we know or don’t know about it. If you carefully search for cases of actual diabetic cardiomyopathy, it appears to be difficult patients to find. Nevertheless, it is important that we look for shared risk factors such as hypertension – and then metabolic derangement on top of those factors.
Is Heart Failure Treatment Different in Patients With Diabetes?
Michel Komajda, MD (University Pierre et Marie Curie, Paris, France)
Former ESC President Dr. Michel Komajda sought to answer an important question: are all diabetes drugs safe in patients with heart failure? He provided a class-by-class overview.
- Metformin appears to be safe and perhaps slightly protective.
- Sulfonylureas (especially first generation SFUs) lean towards slight harm, although much of the existing data is largely not from prospective randomized studies and therefore not definitive.
- Insulin suffers from a major scarcity of data on heart failure. Some studies have shown bad outcomes in insulinized patients with heart failure, but there is significant room for bias given that insulin-treated patients are generally older and have more advanced diabetes. In ORIGIN, a trend towards a reduction in hospitalization for heart failure was seen with Lantus (insulin glargine), although patients in the trial were better treated from a cardiovascular risk standpoint (i.e.: with statins) than in older outcomes studies).
- TZDs have long been known to have a class effect on heart failure, although the increase in risk appears to be lower for Takeda’s Actos (pioglitazone) than GSK’s Avandia (rosiglitazone). One reassuring piece of data was that in the PROactive trial, there was an increase in hospitalization for heart failure but not as much of an increase in the risk of fatality due to heart failure.
- DPP-4 inhibitors were cast into the heart failure debate when the SAVOR CVOT for AZ’s Onglyza (saxagliptin) found a 27% statistically significant increase in hospitalization for heart failure, which Dr. Komajda characterized as surprising and without a known mechanistically plausible explanation. If anything, Dr. Komajda pointed out, one would have expected a decrease in heart failure due to DPP-4 inhibitors’ beneficial modulation of protective BNP. Beyond presenting existing clinical data data, Dr. Komajda did not make any direct speculations about whether the theorized DPP-4 inhibitor class effect on heart failure is likely to be true.
- GLP-1 agonists have evidence of cardioprotective effects in preclinical studies, and the blood pressure and endothelial function benefits suggest the possibility of a positive effect in heart failure. However, the class’ increase in heart rate could counteract these benefits.
Questions and Answers
Q: What do you think about SGLT-2 inhibitors? It looks to be an interesting class – perhaps they will show a benefit?
A: Based on their mechanism of action, there is no reason to anticipate any cardiovascular safety issues. Of course, we are in an era of evidence-based medicine, and so we will need to wait to get the data to say for sure. The example of DPP-4 inhibitors exemplifies the need for randomized control trials. Nobody would have expected that sort of harmful heart failure signal. I think you’re right, since they are purely acting in the kidney, a priori there is no reason to think that they will be harmful.
Q: You mentioned the discrepancy between the results from ORIGIN and previous trials. Could that be because there were improvements in glycemic control in ORIGIN that could have had some positive effect?
A: I think that is one explanation. The other is that the population tested in ORIGIN was different than those exposed to insulin more broadly because the trial enrolled rather recently diagnosed diabetes patients.
Q: Would you advocate to the audience that they should not use sulfonylureas in patients with heart failure?
A: I am not sure that I would go that far, because these were observational studies with obvious limitations, but the data does merit a warning. I would be a little cautious in patients with severely compromised ejection fractions.
Is Heart Failure Treatment Different With Diabetes Patients?
Krishna Prasad, MD (St. Thomas’ Hospital, London, UK)
During his presentation, Dr. Krishna Prasad drew the audience’s attention to a notable secondary finding from the PARADIGM-HF trial, which tested Novartis’ new heart failure drug LCZ696. The results, which were presented earlier at ESC, showed that the drug led to 20% reduction in death due to cardiovascular causes. Dr. Prasad pointed out that the results were slightly less positive for the 35% of trial participants that were also diabetes patients. Although both the diabetes patients and individuals without diabetes saw a marked improvement in the study’s primary endpoint, a composite of CV death and hospitalization for heart failure, the point estimate for the reduction of cardiovascular death alone in diabetes patients was much more modest than the point estimate for nondiabetic individuals (~10% reduction [NS] vs. ~25%, respectively, estimated from a graph). This finding, if true, could be seen as support for previous findings that diabetes patients experience worse outcomes following adverse CV events.
-- by Manu Venkat and Kelly Close