August 29-September 2, 2015; London, UK; Full Report – Draft

Executive Highlights

The European Society of Cardiology Congress 2015 (ESC 2015), just held in London late last week, featured new data and commentary on the latest CVOT results, a significant focus on prevention, and a great deal of excitement around the new PCSK9 inhibitor class for LDL lowering. Below we include our top five highlights from the meeting, followed by full in-depth coverage.

1. Dr. Frans Van De Werf (University of Leuven, Leuven, Belgium) presented a new analysis of TECOS (CVOT for Merck’s Januvia [sitagliptin]) showing absolutely neutral heart failure results across all prespecified subgroups and heart failure-related outcomes. Wow what a lot of grief the class got for what seems essentially now to be considered no big deal.

2. There was significant discussion throughout the conference of potential explanations for the divergent heart failure results in the three DPP-4 inhibitor trials, with a chance finding in one of the trials or distinct drug-specific effects granted the most credence. What bad luck.

3. Several speakers offered broader commentary on the CVOT requirement for diabetes drugs, with the renowned Dr. Robert Califf (Duke University, Durham, NC) who is now at the FDA arguing (somewhat) in favor and Dr. Jaakko Tuomilehto (University of Helsinki, Finland) and Dr. Philippe Gabriel Steg (Université Paris-Diderot, Paris, France) arguing strongly against.

4. Several prevention-focused sessions emphasized the importance of controlling blood pressure and lipids relative to glucose in patients with diabetes and featured a series of debates on obesity, including over the importance of visceral adipose tissue and the existence of the “metabolically healthy obese.”

5. Sessions on the new PCSK9 inhibitor class for LDL lowering consistently drew packed crowds, as speakers highlighted the impressive progress over the past decade and the high potential for cardiovascular risk reduction.

 

Diabetes Drugs

Hot Line III: Diabetes Mellitus/Pharmacology

Trial Evaluating Cardiovascular Outcomes with Sitagliptin in Patients with Type 2 Diabetes: TECOS

Frans Van de Werf, MD, PhD (University of Leuven, Belgium)

In the biggest diabetes-related highlight of the conference (maybe tied for the CVOT session), Dr. Frans Van de Werf (University of Leuven, Leuven, Belgium) presented a new analysis of TECOS (CVOT for Merck’s Januvia [sitagliptin]) showing absolutely neutral heart failure results across the board. The analysis confirmed that the overall neutral results presented at ADA applied across all prespecified subgroups, notably including prior history of heart failure (as well as age, sex, BMI, diabetes duration, A1c, insulin treatment, renal function, prior CAD, and heart failure severity). Results were also neutral for all heart failure-related outcomes measured, including hospitalization for heart failure or CV death, hospitalization for heart failure or all-cause death, total hospitalization for heart failure events, and mortality among patients hospitalized for heart failure. While the reason for the divergent heart failure results between the three DPP-4 inhibitor trials remains a matter of at least theoretical debate, this data should provide welcome additional reassurance on Januvia’s safety and help rule out the hypothesis of a class effect.

• Baseline characteristics varied slightly, in a predictable manner, between participants who were hospitalized for heart failure (n=457) and those who were not (n=14,214). Patients hospitalized for heart failure were older (mean age 68.5 vs. 65.4 years), had worse renal function, and were more likely to have had prior vascular disease or a prior history of heart failure. They were also more likely to be treated with insulin and a diuretic, though both groups had high rates of medication use for diabetes and other cardiovascular risk factors.
• As presented at ADA, the overall hazard ratio for time to first heart failure hospitalization was a right-on-the-nose 1.00 (95% CI: 0.84-1.20; p=0.95). The Kaplan-Meier curves were essentially superimposable throughout the duration of the trial.
• The neutral results were consistent across all prespecified subgroups. Dr. Van de Werf particularly highlighted the fact that results remained neutral when patients were stratified based on prior history of heart failure, as this was the subgroup with the greatest increased risk in SAVOR. The one subgroup that statistically favored placebo (lower bound of the 95% confidence interval slightly above 1) was the group of 45 patients with an unknown New York Heart Association (NYHA) heart failure class. However, we are inclined to agree with Dr. Darren McGuire (University of Texas Southwestern Medical Center, Dallas, TX), who noted during his talk later in the day that the wide confidence interval, very small number of patients, and lack of clarity around the diagnosis cast plenty of doubt on this finding. In addition, the overall interaction for NYHA heart failure class was non-significant.
• All heart failure-related outcomes were similarly neutral. The analysis included results for hospitalization for heart failure (unadjusted, adjusted for baseline heart failure, and multivariate adjusted), hospitalization for heart failure or cardiovascular death (unadjusted and adjusted for baseline heart failure), hospitalization for heart failure or all-cause death, total hospitalization for heart failure events (as well as the subgroups of patients who had two events and three or more events), and mortality (cardiovascular and all-cause) among patients hospitalized for heart failure. Results were resoundingly neutral in all categories.
• All heart failure-related outcomes were neutral among the subgroup of patients with a prior history of heart failure. While overall event rates were significantly higher in this subgroup, there was no difference between the Januvia and placebo groups in rates of hospitalization for heart failure, CV death, a composite of the two, and all-cause death.
• Dr. Van de Werf suggested that intrinsic drug-specific effects or chance were both plausible explanations for the divergent results in the three DPP-4 inhibitor trials. He expressed some skepticism about other potential explanations, noting that SAVOR and TECOS enrolled a similar patient population, that patients were well treated in all three trials, and that heart failure was similarly defined and adjudicated in all the studies. He did not speculate about the implications of the results for the DPP-4 inhibitor class as a whole but stressed that it is clear that Merck’s Januvia can be safely used without an increased risk of heart failure.

TECOS: Discussant Review

Jaakko Tuomilehto, MD, PhD (University of Helsinki, Finland)

Dr. Jaakko Tuomilehto addressed the difficult question of why the three DPP-4 inhibitor CVOTs that have reported results had such different findings with regard to heart failure. He noted that in SAVOR, patients with the highest levels of NT pro-BNP (a biomarker of heart failure) at baseline had the greatest increase in heart failure risk. This remains one of the few unanswered heart failure questions from TECOS, as BNP levels have not yet been analyzed, but we would be surprised if anything significant arose given how consistently neutral the results have been thus far. Dr. Tuomilehto agreed with Dr. Van de Werf’s list of possible explanations for the divergent results but stated that he could not offer any additional insight. He did express skepticism about the feasibility of designing trials specifically intended to evaluate a drug’s effect on heart failure given the relatively low incidence of heart failure compared to total mortality and MI. However, we have heard others, including EXAMINE investigator Dr. Faiez Zannad, suggest that heart failure could be a component of the primary composite endpoint in these trials given its high relevance for patients with diabetes, which seems logical to us.

Dr. Philippe Gabriel Steg (Université Paris-Diderot, Paris, France): Can you comment on whether we need to do randomized controlled trials to measure the safety of drugs that are so widely used when the incidence of adverse events is so low? If we continue down that route, we could spend the entire resources of all clinical research on safety studies that should be the work of pharmacovigilance.

Dr. Van de Werf: Good question. The FDA asked us to do these trials to exclude any potential negative impact on CV outcomes.

The Evaluation of Lixisenatide in Acute Coronary Syndromes (ELIXA)

Eldrin Lewis, MD (Brigham & Women’s Hospital, Boston, MA)

Dr. Eldrin Lewis reviewed the neutral primary results of ELIXA (CVOT for Sanofi’s Lyxumia [lixisenatide]) presented at ADA, with a particular emphasis on the neutral heart failure results due to the controversy surrounding the DPP-4 inhibitor class. While he, unlike Dr. Philippe Gabriel Steg in his subsequent presentation, did not comment on whether he feels the neutral results entirely rule out the potential for cardiovascular benefit, he did outline the “well-established” rationale for cardioprotection with the GLP-1 agonist class as a whole: improved endothelial function, decreased inflammation, weight loss, and favorable changes in lipids and blood pressure.

The ELIXA Trial Results: What Do They Mean?

Philippe Gabriel Steg, MD (Université Paris-Diderot, Paris, France)

Dr. Philippe Gabriel Steg offered his take on the reassuring ELIXA results along with fairly strongly worded commentary on the broader CVOT requirement for diabetes drugs. He emphasized ELIXA’s importance in assuring safety in patients “at the extreme end of the high-risk spectrum” and particularly of the neutral heart failure findings in the context of the ongoing DPP-4 inhibitor controversy. Notably, he also argued that the neutral results do not rule out the possibility of a cardiovascular benefit with Lyxumia (or presumably other GLP-1 agonists), as the trial had a relatively short duration and enrolled high-risk patients with longstanding diabetes who were not representative of the general diabetes population. These limitations apply to ongoing CVOTs of other GLP-1 agonists as well, although Novo Nordisk has repeatedly expressed optimism that the LEADER trial for Victoza (liraglutide) should stand a greater chance of demonstrating superiority due to higher total patient exposure. Dr. Steg then pivoted to a look at the 2008 CV Guidance (and comparable EMA regulations) more broadly, noting that one “has to wonder if this is not a waste of resources” that could be used for more important tasks. As another unfortunate consequence, he also suggested that non-specialists in particular have mistakenly interpreted neutral results from non-inferiority outcomes trials to mean that the drug is not efficacious. We heard similar concerns from the highly regarded Dr. Allison Goldfine (Joslin Diabetes Center, Boston, MA) at Keystone and wonder how widespread this problem is. While Dr. Steg acknowledged the positive topline results from EMPA-REG OUTCOME (which could be used by proponents to justify the value of the CVOT requirement – though we’d much rather see organizations themselves decide on long-term trials rather than having the trials limit the safety characterizations), he concluded by suggesting that “it is time to stop wasting resources on non-inferiority trials and focus resources on truly improving diabetes care.” Hear, hear.  

Lessons Learned from the TECOS Cardiovascular Safety Trial (Sponsored by Merck)

The Role of Sitagliptin in the Treatment of Cardiovascular Patients with Type 2 Diabetes

Eric Peterson, MD (Duke University, Durham, NC)

Dr. Eric Peterson reviewed the primary results from TECOS presented at ADA and emphasized their reassuring nature. He argued that the three DPP-4 inhibitor CVOTs taken together send a very positive message about the class and that TECOS in particular was reassuring with regard to heart failure. He also focused on the positive results for pancreatic safety, showing no significant difference in event rates between groups and very low rates of pancreatitis or pancreatic cancer overall. Dr. Peterson devoted the majority of his allotted time to Q&A, and it was clear from the discussion that heart failure remained top of mind for the attendees. Dr. Peterson provided a few tantalizing hints regarding the in-depth heart failure results presented the following day, describing them as about as reassuring as possible. Like the rest of the field, he could offer no clear explanation for the divergent heart failure results in the three DPP-4 inhibitor trials, joking that “if I had the complete answer, you could hand me a Nobel Prize.” He listed different patient populations, different definitions of heart failure, drug-specific effects, and other background medications as possible explanations but offered reason to be skeptical about all of them. He suggested that statistical chance must be considered as a possibility but expressed confidence in the robust nature of TECOS in particular. One attendee inquired about the recent FDA communication related to joint pain with DPP-4 inhibitors; this concern appeared to be on very few people’s radar before the FDA warning, and we believe it is likely to become more of an issue going forward though it was a small number of complaints. Dr. Peterson did not recall the specific rates of joint pain in TECOS but stressed that there were no imbalances in any secondary endpoints. Finally, we were very eager to listen when the discussion turned to the utility of longer studies designed to assess the effects of glucose lowering achieved with specific medications rather than these shorter trials intended primarily to demonstrate safety – we would love to see the field move in this direction.

Questions and Answers

Q: You didn’t tell us about the percent reduction in A1c. The patients started with A1c near controlled, which is really not where we would be prescribing this.

A: The sitagliptin arm had a significantly lower A1c, but this was about the range people wanted to titrate to, and over three years there was no increase like you often see in type 2 diabetes, so that was positive. There was also no need to add in insulin as quickly in the sitagliptin group.

Q: Was the risk of heart failure different in the 17% with a baseline history?

A: Tomorrow we will present the results of a late-breaking trial that goes into the in-depth results on the heart failure population. We’re also working on a manuscript that will give those results. In the heart failure subgroup, among those with prior heart failure, there was no signal for increased risk.

Q: For the elevation in risk for saxagliptin, is that hospitalization for heart failure or incident heart failure?

A: The definition was hospitalization for heart failure across all three trials. They didn’t distinguish between recurrence or first onset. In tomorrow’s presentation they will go over the results for the first incident case and development of subsequent hospitalization.

Q: Does the difference in selectivity for DPP-4 relate to the difference of these trials in heart failure?

A: If I was here with that complete answer, you could hand me a Nobel Prize. There are several potential opportunities of why that might be. There might be different patients enrolled. EXAMINE looked at ACS patients, SAVOR had a secondary population and high-risk primary prevention, and TECOS just had people with secondary disease. Across the studies, there are more similarities between SAVOR and TECOS than there are differences. To my eye, although you can’t rule out differences in patient populations, it seem the populations were more similar than different. If I look within subgroups as we will, no subgroup has a signal at all. Prior heart failure, renal problems, longer-term diabetes – you can keep going down the list, but there’s no subgroup with increased risk. It’s remarkably reassuring. I don’t think it was explained by baseline characteristics. Could it be different definitions of heart failure? They used very similar definitions across the three trials. You can’t rule out how the cases were defined, but on paper it was the same. Was it a difference in the drugs themselves? Possibly. There are biologic differences between agents and various explanations for why in retrospect they may differ. Those are post-hoc theories. We have to wait for a larger study or a comparative study to define that. Could it be due to other drugs? Unlikely. The use of drugs was very similar in the populations and if you take subgroups and look at those on RAS inhibitors or diuretics, there was no difference. Statistical chance has to be mentioned, but in TECOS, there were over 14,000 patients followed for three years and there was no safety signal. I can’t tell you. I can tell you that sitagliptin looks remarkably stable.

Q: The FDA reported 28 cases of severe joint pain with sitagliptin. Did you look at that in TECOS?

A: We have a series of investigator-reported events, but it wasn’t a primary endpoint that was adjudicated. I don’t know the answer. We looked at a huge number of non-cardiovascular endpoints and in no groups did we see an imbalance.

Q: In the publication it says 80% of patients were still on sitagliptin when the trial ended. Did you look into differences in outcomes for those who took the drug vs. those who didn’t?

A: Very few people were lost to follow-up. The rates of discontinuation were quite similar between the groups and if anything higher in the placebo group. Is there any difference in the results? The short answer is that the on-treatment analysis looks identical to the intention to treat analysis. For the primary endpoint, they were equivalent.

Q: I know this was a response to FDA requirements and it was self-reported exposure; they reported how much they have taken. Was there any request for PK sampling or a measurement that they were taking the drug?

A: It was not required. It was self-reported and investigator-reported. Patients were followed closely, but no. We do have blood samples, so we could go back and confirm on a subset of patients.

Q: Sometimes patients have symptoms of heart failure that can be managed as an outpatient. Was this documented?

A: We have tried to look for cases that weren’t labeled as heart failure, and no, it didn’t change the results. There’s always some fuzz around hospital admissions because some heart failure flares could be managed as an outpatient. There are also cases of misdiagnosis and cases that are diagnosed as heart failure that aren’t. We’ve done the best we can. It was a blinded adjudication, the same as in the other studies, and there was no sign.

Q: TECOS, EXAMINE, and SAVOR were all industry-funded trials to fulfill regulatory requirements. Should we be doing more to push for UKPDS-like trials with 20-year follow-up?

A: It’s a different question to ask whether drugs modify long-term outcomes. The design would be very different. It’s not the fault of industry, to be honest. The reason we did the study this way with regard to minimizing A1c differences was to differentiate using a strategy with sitagliptin to get A1c within range as safe vs. alternative strategies. The study confirmed in that setting, when A1c was pretty darn close to target, using sitagliptin is safe. If we designed a different long-term study, if a person is on sitagliptin long-term and A1c is better controlled, is there a long-term difference in outcomes? That would take many years, maybe up to a decade, before we would see a difference based on previous trials. It’s important information. Prior trials have shown that long-term tight control results in better outcomes. But we want and need information on the safety of drugs because they’re widely used. We have part of the answer. The drugs are safe. Now the next level is more studies to show if various strategies to achieve a long-term A1c in range translate to better outcomes. These are great studies to do.

Meet the Trialist: TECOS

Darren McGuire, MD (University of Texas Southwestern Medical Center, Dallas, TX)

Dr. Darren McGuire reviewed the findings from the new TECOS heart failure analysis presented earlier in the day. He described the finding of increased risk in SAVOR (for which he was also an investigator) as “completely surprising” and assured attendees that the TECOS team ensured they were as thorough as possible with regard to assessing heart failure risk following those results. He, like most other speakers at this conference, views the results from EXAMINE as a legitimate safety signal, whereas others have interpreted them as neutral due to the lack of statistical significance. As he went through the various sub-analyses, Dr. McGuire emphasized that “no matter how granularly we slice the data,” there was absolutely no hint of increased risk. As for the question of the hour – how to explain the divergent results from the three trials – he did not have a magic solution, proposing the now familiar list of possibilities. However, he did stress that if one is going to attribute the difference to chance, “you can’t cherry pick,” and argued that it is just as likely that the favorable TECOS results were due to chance as the surprising findings from SAVOR. The point was an important reminder of this possibility, although we have heard many other speakers suggest that as TECOS was the most robust trial of the three, there is reason to give those results disproportionate weight.

Questions and Answers

Dr. Jaakko Tuomilehto (University of Helsinki, Finland): There is one significant finding at the bottom of the slide for one group divided by heart failure severity. Can you explain that?

Dr. Darren McGuire (University of Texas Southwestern Medical Center, Dallas, TX): I would caution against interpreting that as a true finding. The interaction for heart failure severity was not significant. There were only 45 events in that group with undetermined severity of HF at baseline – this was the group where the NYHA class couldn’t be determined, and you see the wide confidence interval. This was probably play of chance.

Dr. Frans Van de Werf (University of Leuven, Belgium): You have to look at the categories together. There is only one p-value and it was not significant.

Dr. Tuomilehto: I agree it was probably a chance finding.

Q: Mechanistically, how do DPP-4 inhibitors increase heart failure? What is the difference between agents?

Dr. McGuire: The short answer is there’s absolutely no plausible explanation put forward to date. DPP-4 was thought to be a unique degrading enzyme of GLP-1 and GIP, but we’ve now characterized some 35 substrates in the blood. One substrate is BNP. The biology there suggests it should actually reduce heart failure. The best basic science biology told us to anticipate favorable effects, and now a large-scale randomized trial saw a signal in the opposite direction. There’s Substance P, PYY, and others that are modified, so there may be explanations. We also don’t know if this was systolic or diastolic dysfunction – there’s a long list of things. We do know that BNP is, if anything, flat, and in EXAMINE it declined in parallel. There was no difference between groups. In SAVOR, BNP at baseline predicted heart failure but there was no differential effect of saxagliptin based on that. There was no clear BNP signal and no clear weight gain. With the TZDs we learned peripheral edema is often a signal, but there was no increase in peripheral edema with any of these three.

We don’t have a clear explanation. In all three, heart failure was prospectively collected and adjudicated in the same way. It may be differential treatment within the class. It may be play of chance. We want to say SAVOR was a play of chance but there’s a similar probability that ours is chance. We can’t cherry pick. Ours tells us what we want to hear. Vildagliptin has been associated with reduced end diastolic volumes. There are two ongoing trials with linagliptin, so we’ll have another trial to contribute.

Q: Were the baseline characteristics exactly the same in SAVOR and TECOS?

Dr. McGuire: There were differences in study eligibility criteria. But if you put head to head the baseline characteristics of the three, they were quite similar. In SAVOR, to be eligible you had to have type 2 diabetes with either atherosclerotic vascular disease or a clustering of CV risk factors. The SAVOR trial did have a slightly higher event rate; it’s unclear why. It may have to do with background therapy or how advanced the disease was. SAVOR allowed worse control of diabetes at study entry, with an A1c up to 10% instead of the 8% upper limit in TECOS. EXAMINE was post-ACS patients. In TECOS, people weren’t excluded with ACS but the bulk of patients had chronic stable CVD.

Dr. Van de Werf: The rate of hospitalization for heart failure was about the same in SAVOR and TECOS. It was higher in EXAMINE, which was explained by ACS.

Dr. McGuire: The annualized event rate for hospitalization for heart failure was 1.2% in SAVOR and 1.1% in TECOS. That’s almost identical risk. You can’t find anything other than the drug that’s different.

Dr. Tuomilehto: Maybe once they knew BNP [in SAVOR], they looked more carefully at patients and made more diagnoses.

Dr. McGuire: The reality is that in SAVOR, BNP was analyzed in batch assays. It was not available to investigators in real time. We don’t have BNP measured in TECOS yet. There was much debate on whether to publish the HF subanalyses data from TECOS without the BNP results, or wait on them. Our biobank assays will be confounded by decision-making and politics about how best to use those precious samples. We didn’t want to wait to get the data out. Because inhibiting DPP-4 may alter BNP in a way that isn’t reflected in commercial assays, we may also be overestimating circulating BNP. But that’s pure speculation.

Dr. Tuomilehto: How would you resolve the issue now?

Dr. Van de Werf: If you look at the meta-analysis, from a pure statistical point of view the p-value is negative and there’s a 14% increased risk. There is a test of heterogeneity that suggests differences between the trials. I’m not going to prescribe DPP-4 inhibitors because I’m a cardiologist, but if you’re concerned about heart failure risk for a particular patient, you should probably select sitagliptin because the data is more reassuring. That’s a simple point of view, but there’s some good data to support it.

Dr. McGuire: Short of a head-to-head trial, which we know will never be done, we’re seeing a number of sophisticated pharmacosurveillance studies. There’s some potential to do sophisticated analysis. We’re spending this much money over and over to find neutral results, so we may need better ways to do this without randomized trials. We could follow up on the SAVOR cohort to see if long-term outcomes track with the incremental heart failure hospitalization risk. We know there was no other signal for adversity. The composite of heart failure with death was not increased. We don’t know if it’s real and it’s hospitalization for heart failure, but is that heart failure like we know it or a little volume overload that goes away when you stop the drug?

Dr. Tuomilehto: In SAVOR, the highest BNP quartile had a 20% incidence of heart failure, which is quite high. It could be possible to design a trial only using patients like that to see what happens. A head-to-head trial with different DPP-4 inhibitors may not be possible but it may not be impossible either.

Dr. McGuire: I’ve heard a proposal for using baseline BNP at screening. Even though the highest quartile had the highest absolute incidence, that was true in both arms. The relative incremental risk was comparable. It was a 2% absolute risk and a 30% relative increased risk. The absolute risk gets to the point that it might inform decision-making. Otherwise slight elevations aren’t informative for clinical prescription choices.

Dr. McGuire: Endocrinologists in the audience, would you prescribe a DPP-4 inhibitor to a patient at risk of heart failure?

Q: I’m a cardiologist and a diabetologist and we use a lot of DPP-4 inhibitors. They’re very good. What about neuropeptide epsilon, another substrate of DPP-4?

Dr. McGuire: That’s the first I’ve heard of that peptide. Maybe you can teach us.

Q: There’s a possibility that could worsen heart failure.

Dr. Tuomilehto: So far I’m not aware that the EMA or FDA would have warning statements about DPP-4 inhibitors and heart failure, are you?

Dr. McGuire: The FDA convened an entire session to review SAVOR and EXAMINE. As I understand it, the decision seemed likely that there would be a caution for saxagliptin and alogliptin wouldn’t need it. The FDA pivoted on the statistical significance issue more than the point estimates. It looks like alogliptin will get by and saxagliptin will have a label change required. Not a contraindication or boxed warning but a caution. I’m confident sitagliptin will get a pass too.

Dr. Van de Werf: I’m not aware of anything from the EMA. So far there’s been no warning.

Dr. Tuomilehto: Hopefully AstraZeneca will need to do additional studies.

Dr. Tuomilehto: If you designed TECOS again, would you have a baseline assessment of BNP?

Dr. Van de Werf: Yes. What about imaging? That wasn’t available. But if you incorporate imaging in a huge trial, the cost increases exponentially, and it’s just capturing one moment in the course of the trial. So that’s not feasible.

Dr. McGuire: I am involved in another ongoing trial of a DPP-4 inhibitor where ejection fraction will be embedded into the data collection if available in the clinical record. At least in the US, virtually all patients will have ejection fraction estimated in the hospital when admitted for heart failure. What we failed to do in TECOS and SAVOR is capture that prospectively, but hopefully we will correct that for ongoing and future DPP-4 inhibitor trials. In TECOS I don’t think it would be that important, but I sure wish we had imaging in SAVOR.

Q: 40% of patients had a previous history of heart failure. Were you able to look at the cause? Do you think differences in the cause of heart failure could translate into different results?

Dr. McGuire: We didn’t capture the etiology of heart failure at baseline in TECOS. We were as surprised as anyone when the signal emerged in SAVOR. I wish we had captured more granular data. TECOS had already enrolled completely, so we could only play catch-up. All we have is a yes or no and the NYHA classification.

Should the Same Rules Apply for Drug and Device Approval? EU vs. US Perspective

Hans Hillege, PhD (University of Groningen, Netherlands) Robert Califf, MD (Duke University, Durham, NC), Eric Klasen (Medtronic, Tolochenaz, Switzerland), Robert Scott, MD (Amgen, Thousand Oaks, CA) Christian Hamm, MD, PhD (Alfred Wegener Institute, Bremerhaven, Germany) Steven Nissen, MD (Cleveland Clinic, Cleveland, OH) Alan Fraser, MB, ChB (Cardiff University, Cardiff, UK), and Christopher Cannon, MD (Brigham and Women’s Hospital, Boston, MA)

This series of mini-presentations and discussions revealed a range of opinions about the different approaches taken by regulatory authorities in the US and the EU, particularly with regard to medical devices. For drugs, most participants agreed that the approaches in the two regions are based on the same set of principles and are largely reasonable. There was also a general consensus that there are enough differences between drugs that they should should not be lumped together under the same rules, though not everyone agreed. Several speakers felt strongly that the current EU system for approving medical devices is inadequate and should move closer to the US approach, though some touted the earlier access to new devices in the EU as an important advantage. Beyond these general threads, the discussion also touched upon a wide range of relevant topics, including the 2008 CV Guidance for diabetes drugs, the regulatory climate for obesity drugs in Europe, and the flaws of the current FDA Advisory Committee system. Below we include several of the most notable quotes from the session.

• Dr. Califf on CVOTs for diabetes drugs: “People should pay attention to the recent press release about empagliflozin. The one that people thought was the least likely to affect outcomes has done it. A trial designed to show safety is the same size as one designed to show efficacy. What we want are drugs with beneficial effects on outcomes. We should do appropriate clinical trials, and we need to improve the way we do them so we’re not wasting money.”
• Dr. Hillege on obesity drugs in Europe: “There’s some historical experience from Europe because starting in 2010 we had to remove two drugs from the market, so at the CHMP level people are really careful. Now things have changed. It’s similar to the FDA’s decision-making, almost on the same level…I agree overall it’s currently different but the FDA and EMA are moving forward in harmony.”
• Dr. Scott on FDA Advisory Committees: “The FDA’s use of expert committees for drugs is very painful from an industry perspective. In recent years it has been difficult for the FDA to appoint members because the rules around conflicts of interest means most prominent thought leaders in a given field are conflicted. The meetings also require viewing a lot of materials, so many don’t come to the committee fully briefed. From an industry perspective, you expend a huge amount of resources to prepare and the results are somewhat arbitrary and random. The FDA has issues with these meetings too. One division director referred to them as a pack of hyenas circling a wounded carcass.”

Managing Diabetes & CVD: Exploring New Evidence and Opportunities (Supported by BI/Lilly)

SGLT-2 Inhibition in the Management of Type 2 Diabetes: Potential Impact on CVD Risk

Silvio Inzucchi, MD (Yale University, New Haven, CT)

Despite the tantalizing title, Dr. Silvio Inzucchi’s presentation did not touch upon the recently announced positive topline results from the EMPA-REG OUTCOME study of Lilly/BI’s Jardiance (empagliflozin). However, he did review the reasons why so many have held out hope that the SGLT-2 inhibitor class could have a positive effect on cardiovascular outcomes: blood pressure reductions, improvements in albuminuria, glucose and insulin reductions, reductions in weight and visceral adiposity, less oxidative stress, and some (non-LDL) beneficial lipid effects. He also emphasized the importance of the low hypoglycemia risk seen with SGLT-2 inhibitors and other newer diabetes drug classes, describing hypoglycemia as the “rate-limiting factor” in diabetes care for many clinicians. Dr. Inzucchi also referred to the ongoing concerns about DKA with SGLT-2 inhibitors, though he characterized the risk as primarily a concern with off-label use in type 1 diabetes that is directly related to insulin deficiency. We will be very curious to see the rates of DKA in EMPA-REG OUTCOME and the other large outcomes trials of SGLT-2 inhibitors to better understand the risk in type 2 diabetes, but all evidence thus far indicates that it is quite rare in this population.

SGLT-2 Inhibitor Outcome Trials: Future Opportunities

David Fitchett ,MD (University of Toronto, Ontario, Canada)

Dr. David Fitchett danced even more tantalizingly close to the EMPA-REG OUTCOME results but warned the audience from the start that he planned to “leave you in suspended animation” until full results are presented at EASD in a matter of weeks. He provided a succinct summary of the status quo with regard to diabetes and cardiovascular disease: diabetes causes increased risk, the benefits of glucose lowering are uncertain, there may be harm from over-treatment (primarily based on the increased mortality in ACCORD), and it is possible that treating early with specific agents could be beneficial. He did cast doubt on the common assertion that metformin leads to cardiovascular benefit, arguing that its first-line status is based on a “grossly underpowered” sub-study of UKPDS – it would certainly be informative to see a dedicated outcomes trial investigating the long-term effects of metformin (as well as other drugs approved before the 2008 CV Guidance). While he repeatedly cautioned against over-interpreting meta-analyses, Dr. Fitchett did present data suggesting the potential for cardiovascular benefit with SGLT-2 inhibitors. Interestingly, he described selectivity for SGLT-2 vs. SGLT-1 as one of the few differences between the three approved agents and suggested that this could be put forward as an explanation if results from the three outcomes trials are different. This was an interesting hypothesis, though our sense is that for what is likely a subtle effect, differences in trial design are at least as likely an explanation as actual drug-specific effects for any divergent results.

Medication in Cardiovascular Patients: The Assumed Risk

Antidiabetic Drugs and Cardiovascular Risk

Eberhard Standl, MD (Helmholtz Centre, Munich, Germany)

Dr. Eberhard Standl used a case study (61 year old man with uncontrolled type 2 diabetes, mild microvascular complications, and a recent MI) as a framework to discuss considerations involved in setting A1c targets and selecting appropriate medications for high-risk patients with type 2 diabetes. He argued in favor of an A1c target at the higher end of the spectrum (as high as 8% initially) for such patients based on current guidelines and emphasized the enormous risks of severe hypoglycemia. He suggested that it would be preferable to initially choose “non-insulinotropic agents” to avoid excess risk of hypoglycemia – we agree, though we do take some issue with his inclusion of GLP-1 agonists in the less preferable category, as the class has been shown to carry a very low risk of hypoglycemia. Dr. Standl then briefly reviewed the available evidence on the CV effects of the various drug classes. He described the topline EMPA-REG OUTCOME results as very promising and suggested they could potentially lead to a paradigm shift in type 2 diabetes treatment, though he expressed some concern about the increases in LDL, the risk of DKA, and other side effects of SGLT-2 inhibitors. On the DPP-4 inhibitor/heart failure question (a seemingly ever-present focus at this conference), Dr. Standl did not offer one clear explanation but seemed to grant plausibility to both a chance finding in SAVOR or divergent drug-specific effects.

Questions and Answers

Dr. Dan Gaita: As cardiologists, we’re always looking for CV protection, though we’re not enthusiastic that microvascular problems are not well controlled either. Sitagliptin is safe. Pioglitazone is still alive but there are all these rumors. What do you recommend about that?

Dr. Standl: Looking back from today, if the investigators just went for the outcome of MACE, the results [from ProACTIVE] would have been highly positive. Those people trusted the vascular surgeons and also included revascularization of the lower limbs and I think screwed it up. There are some reservations. Heart failure is a contraindication, but it’s a strong oral drug for lowering fasting glucose. It has merits, and it can be used down to a low GFR. It has a place, but with newer compounds, especially if it’s true what has been promised for SGLT-2 inhibitors, there are alternate options.

Dr. Sattar: This patient is very high-risk; he’s had diabetes for almost a decade and existing CVD. He has microvascular disease as well. Based on a recent paper in JAMA, he has a reduced life expectancy of 10-15 years. Reducing CV risk and mortality is important. I would tend to reduce blood pressure more and increase statin intensity. My thought is we should be avoiding sulfonylureas in this type of patient with prior severe hypoglycemia. DPP-4 inhibitors reduce sugar safely. For the SGLT-2 inhibitors, is the data going to be a paradigm shifter? Insulin is the last thing we should use. So I would use a DPP-4 inhibitor, see what the SGLT-2 inhibitor data shows, and consider GLP-1 agonists as options.

Q: If his A1c is still above 8% with medications, what further regimen would you pursue?

Dr. Standl: What we actually did was stop the sulfonylurea and switch to a DPP-4 inhibitor, sitagliptin. The problem may be a difference in efficacy. If you remove the sulfonylurea at an A1c value of 8.3% as a starting point, you may get stuck with the comparable lowering of A1c being less striking than you had hoped. I can’t say if he succeeded in getting the A1c down to 7%, which would also involve education and improving lifestyle and all that. This is a limitation if you shift from a sulfonylurea to a DPP-4 inhibitor. DPP-4 inhibitors clearly lower postprandial glucose, but they’re not great for fasting. It’s important to take the metabolic phenotype into account, fasting vs. postprandial hyperglycemia. It helps further differentiate the treatments.

Q: What are your thoughts on the SGLT-2 inhibitors?

Dr. Standl: They are the latest player. There are potential downsides: in particular, LDL may go up, there’s a small diuretic effect, and a problem with genital infections. There are downsides. But it may be a shift in the paradigm if it’s true that you prolong life and reduce CV outcomes. It’s clearly too early to tell. We also don’t know whether it’s true for the whole class or just one compound. It has also been used in type 1 diabetes, where it may mask DKA. That’s something that you have to know if you’re really going into that.

Dr. Sattar: If the relative risk reduction is greater than we imagine, it may well be a paradigm shift. But let’s see the data first. They’re new drugs and they’re more expensive so the data will need to be compelling.

Obesity and Prevention

Mobile Health and Technology: The Future Technical Environment for Prevention

Can Mobile Technology Improve Glycemic Control and Save Costs in Diabetes?

Deborah Chyun, PhD (New York University School of Nursing, New York, NY)

Dr. Deborah Chyun opened her talk by emphasizing the importance of diabetes to the cardiology community and reviewing the mixed evidence on the relationship between glycemic control and cardiovascular outcomes. She suggested that particularly in type 2 diabetes, earlier intervention is likely needed in order for glucose control to have an impact, which speaks to a need for better screening. In the meantime, she advocated for an approach to diabetes management that targets multiple risk factors, though she acknowledged the challenges related to cost and adherence. Turning to the central focus of her talk, Dr. Chyun characterized mobile health interventions as a potentially helpful tool for diabetes self-management education/support but repeatedly stressed the need for better evidence of their effectiveness. She cited several reviews of mHealth studies that identified a long list of problems with their methodology, particularly an inability to rule out confounding factors like familiarity with technology, cognitive ability, and diabetes medications, just to name a few. She also noted that the few studies that have attempted to evaluate the cost-effectiveness of these interventions have not used standard economic analysis techniques. While we continue to believe that digital health holds great potential for diabetes treatment, we appreciate Dr. Chyun’s reality check on just how far the field needs to go before it can become widely accepted. We also think her question of whether randomized controlled trials are the best approach to assess these interventions is a truly important one – as one audience member pointed out, the relatively slow pace of such research may not be compatible with the rate of technological change. To boot, we’re not sure that having patients in trials that can vary so widely from “real life” gives the field the optimal comparisons.

Questions and Answers

Q: In your vast experience with type 2 diabetes, have you found growing interest in finding a way to help people become more behaviorally motivated?

A: Yes. I work a lot with older adults and they’re very interested in using technology to help. We think when people are older they won’t be willing to do it, but it’s not true. One study used a transmission device to transmit glucose directly, which was problematic, but with mobile phones the technology is much easier.

Q: Did adding that to usual practice add a significant amount of time and responsibility?

A: No, they had to transmit a couple of times a week. The initial setup was difficult but then it was fine. People actually enjoyed it. This was seven or eight years ago so the technology was very different.

Comment: That’s the problem. The technology moves so fast, it’s hard for research to keep up.

Reducing Cardiovascular Risk in Diabetes: One Size Does Not Fit All

State of the Art in Reducing Cardiovascular Risk in Diabetes

Naveed Sattar, MD (University of Glasgow, UK)

Dr. Naveed Sattar reviewed the recent progress and remaining challenges in CVD prevention among patients with type 2 diabetes, emphasizing the need for a multifactorial approach. He opened with a refreshingly positive reminder of the significant decline in CV event rates among people with diabetes in the past few decades, which he attributed to earlier diagnosis and more aggressive treatment of blood pressure and lipids. As we have heard repeatedly at this conference, he reminded the audience that treating blood pressure and lipids gives “more bang for your buck” compared to treating glucose in terms of macrovascular outcomes. He described the somewhat disappointing results from trials investigating the relationship between glucose and cardiovascular outcomes as not particularly surprising, as most involved patients with a fairly low baseline A1c, small reductions in glucose over the course of the trial, and good management of other risk factors. In terms of glycemic targets, Dr. Sattar recommended a goal of 6.5% near the time of diagnosis, 7%-7.5% for the majority of patients, and more relaxed targets for select groups like the elderly or those with high hypoglycemia risk or existing CVD. As for the timely question of the effects of specific diabetes drug classes on cardiovascular outcomes, Dr. Sattar expressed some optimism about the potential for GLP-1 agonists and SGLT-2 inhibitors to prove to be beneficial but offered no further specific predictions.  

Improved Cardiovascular Risk Factors Control Associated With a Large-Scale Risk Reduction Program Among Diabetes Population

Jamal Rana, MD, PhD (Kaiser Permanente Northern California, Oakland, CA)

Dr. Jamal Rana presented results from a longitudinal observational study demonstrating significant improvement in CV risk factors among patients with diabetes in the Kaiser Permanente Northern California (KPNC) system after the implementation of a comprehensive prevention program. The PHASE program, first implemented in 2004, aims to deliver evidence-based preventive therapies for blood pressure, lipids, and glucose in a high-risk population with diabetes. The goal was to evaluate changes in risk factor control before and after PHASE implementation and to compare them to national trends, measured by the HEDIS tool used by  >90% of American health plans. Results showed significant improvement in several risk factors from 2004-2013 among the KPNC population that were often greater than those seen nationally. Average LDL levels dropped from 104 mg/dl to 88 mg/dl, and the percentage of patients on a statin rose from 57% to 72%. The percentage of patients with an A1c >9% dropped from 28% to 18% (p<0.05) in the KPNC group while it remained stable at 32% nationally. The percentage of patients with an LDL <100 mg/dl rose from 45-50% to close to 75% in the KPNC group (p<0.05), while it remained stable around 45% nationally. The percentage of patients with blood pressure <140/90 mmHg was not assessed until 2007 and did not show a significant trend between then and 2013, but a higher percentage of KPNC patients (close to 80%) achieved that goal than the national database (~60%). Dr. Rana cautioned that it is impossible to draw causal inferences from this sort of non-randomized study but suggested that the results were promising.

Future Direction for Research in Reducing Cardiovascular Disease Risk In Type 2 Diabetes

Jaakko Tuomilehto, MD, PhD (University of Helsinki, Finland)

His bright red suit providing a dash of color in the sea of black and navy, Dr. Jaakko Tuomilehto argued that effective prevention of CVD in patients with type 2 diabetes will require a shift in the research paradigm to focus on much earlier stages of the disease, likely meaning patients with prediabetes/impaired glucose tolerance (IGT). He noted that the majority of CV events and deaths come not from the population with diagnosed diabetes but from those with IGT, and that the disappointing results from glucose-lowering trials may result from the fact that it is simply too late to affect outcomes by the time diabetes is diagnosed. He highlighted the EU-funded ePREDICE study, which is evaluating the effect of placebo vs. metformin vs. linagliptin vs. metformin+linagliptin, all in addition to lifestyle, on CV outcomes in patients with IGT based on an oral glucose tolerance test (OGTT). Patients will be assessed at 36 months, and Dr. Tuomilehto hopes to extend the study to 60 months if funding permits. He also highlighted the utility of the FINDRISC questionnaire to identify high-risk patients who would be ideal for such studies. Finally, Dr. Tuomilehto offered very strongly negative commentary about the current CVOT paradigm required by the FDA, suggesting that “all the money companies have is going to stupid CV safety studies that do not advance our science and understanding of glucose-lowering therapy at all.”

Questions and Answers

Comment: It’s still a hypothesis that treating borderline diabetes improves outcomes. PPAR gamma drugs should have given incredible returns but they had bad outcomes. We must do trials. What we do know is lowering LDL in prediabetes decreases risk. I hope you’re right, but at the moment it’s a hypothesis.

A: What you’re proposing is to concentrate on lipids rather than glucose. That might be right, but you have to have different kinds of studies. I agree this kind of research is needed, but at the same time you have to understand what happens when you control glucose early on.

Comment: All I’m saying is treating patients for lipids and blood pressure is proven therapy. The hypotheses regarding A1c <7% have been very disappointing. And I wouldn’t say the FDA is stupid; I think they’re being quite sensible.

Q: You discussed use of resources to promote more studies in precision medicine in diabetes, and I think that’s important. On the other hand, one reason why resources are used in a different way is related to the request of regulatory agencies. Should there be a different approach in terms of side effects and safety at the level of the regulatory agencies?

A: Of course it’s important to fully understand how different medicines are actually working and get lots of information. You have to remember when a new drug comes to market after phase 2 and3, the typical case is about 6,000 patients studied, it’s usually a very small number. We need further studies but now the problem is this arbitrary definition of a confidence interval below 1.3, which is completely arbitrary and taken out of the blue. There’s no scientific evidence that this should be the threshold for drugs to show non-inferiority in CV safety. And it means efforts to do further evaluations of new drugs are going that way and not toward looking more broadly at other outcomes and mechanisms. So yes, I agree we should have a broader view of new medications in order to see what they’re actually doing and not doing.

Q: I’m interested in the idea of early screening of diabetes, but around a decade ago we were talking a lot about metabolic syndrome and insulin resistance and we talked a lot about those patients having high CV risk. Why don’t you include in this category of treating early the group that is really high risk with abdominal obesity, etc.?

A: Sure, it could be but we need evidence and studies. There is too little research ongoing at that point in progression. We know dysglycemia is progressive, and the majority of research work is done at the very end of progression, so very little is done at earlier stages. You can have many different innovative research ideas at that level. I agree it can be done this way too.

Managing Diabetes & CVD: Exploring New Evidence and Opportunities (Supported by BI/Lilly)

Diabetes & Cardiovascular Disease: The Risk Factor to be Challenged

John Eric Deanfield, MD (University College London, UK)

Dr. John Eric Deanfield opened this session with the familiar sobering statistics on the global burden of diabetes and the related increased risk of premature CVD and mortality. He then reviewed data from the Emerging Risk Factor Collaboration’s analysis of 102 prospective studies demonstrating much tighter associations between blood pressure and lipids and cardiovascular outcomes compared to glucose control. As he reminded the audience, this is an area of continuing controversy and frustration, embodied by the disappointing results from studies like ADVANCE. However, it is also a strong argument for a multifactorial approach to diabetes care that focuses on cholesterol and blood pressure in addition to glucose, and we expect the standard of diabetes care to increasingly move toward such a holistic approach.

What is the Story About Diabetes & CVD: Rationale for a Multifactorial Approach

John Betteridge, MD, PhD (University College Hospital, London, UK)

Dr. John Betteridge continued the focus on a holistic approach to diabetes care that involves greater integration between the “rich cardiologists and the poor diabetologists,” as exemplified by the joint ESC/EASD guidelines. He explained that as there are physiological differences in atherosclerosis between patients with diabetes and those without (plaque appears earlier and is more aggressive in diabetes), a more aggressive lipid-lowering approach may be required in diabetes compared to the general population. This is one reason why we see patients with diabetes as a population that stands to benefit so much from PCSK9 inhibitors – it is such a big population and of course, more likely than otherwise to be at higher than normal risk of heart disease. Like Dr. Deanfield, Dr. Betteridge emphasized the critical role of blood pressure lowering as well and expressed frustration at the disappointing results from trials investigating the relationship between glucose control and macrovascular outcomes. He explained that he personally believes glucose is related to cardiovascular disease and noted the positive long-term results in type 1 diabetes trials like DCCT/EDIC and UKPDS. However, the results in type 2 diabetes remain unclear, and he emphasized the need to individualize therapy and avoid the clearly negative effects of hypoglycemia.

How Are We Managing Cardiovascular Risk in the Primary Care Environment? EUROASPIRE IV Survey

New ChAllenges in a Changing Research Environment

Kornelia Kotseva, MD, PhD (Imperial College London, UK)

Dr. Kornelia Kotseva outlined the purpose and design of the ESC’s EUROASPIRE IV survey intended to evaluate the implementation of the organization’s cardiovascular prevention guidelines throughout Europe. The objectives of the study were (i) to determine whether the guidelines were followed in everyday practice; (ii) to determine whether implementation had improved since the previous survey in 2006-07; (iii) to compare strategies related to glucose metabolism and chronic kidney disease (which were not included in previous surveys); and (iv) to identify opportunities to improve patient management by revising the guidelines or their implementation. The survey involved exams and interviews of 4,579 patients (out of 6,700 identified through medical records) in primary care centers in 14 European countries; results for coronary patients treated in the hospital were presented in 2013. Participants in this arm had no history of CVD but had been prescribed one or more blood pressure-, lipid-, or glucose-lowering medications at least six months and at most three years prior to the expected date of interview. The outcome measures were the proportion of patients achieving lifestyle, risk factor management, and therapeutic targets for CVD prevention in the current ESC guidelines. As Dr. Kotseva described it, strengths of the survey included its use of standardized methods and interviews and central lab measurements, which should provide a more complete and consistent database compared to many observational studies. The main challenges she identified were low participation in some countries and considerable variation between countries in baseline characteristics.

Questions and Answers

Q: What sort of patient population are we looking at given that you have invited patients who are treated? The proportion treated at a given level of risk may vary across countries. If you have the next round and the proportion has changed, how can we compare?

A: Ideally if we wanted to identify people at high risk of CVD, we should have used a total CVD risk calculator like HeartScore. However, many patients were already on medications and according to the JES Guidelines on CVD prevention, HeartScore cannot be used in people on blood pressure or lipid-lowering medications. We identified patients from the GP practices medical notes. 83% of enrolled patients were on one, 23% on two, and 4% on three medications. We had the same identification criteria in EUROASPIRE III and this gives us the opportunity to compare the two surveys.

Q: But I understand EUROASPIRE III was in specialist care? Would that make a difference in the comparison?

A: EUROASPIRE III and IV had two arms, hospital (coronary patients from hospitals) and primary care (people at high CVD risk from general practices). The results we are reporting today are in the primary care arm of EUROASPIRE IV. We used the same identification criteria for high-risk people in both surveys. We compared EUROASPIRE III and IV in those countries where patients were recruited from the same geographical areas and general practices, although more general could have been added in EUROASPIRE IV. These results will be presented at the Registry/Hot Line session on Tuesday.

Principal Results: Lifestyle and Risk Factor Management

Guy De Backer, MD, PhD (Ghent University, Ghent, Belgium)

Dr. Guy De Backer presented results for the lifestyle and risk factor management components of EUROASPIRE IV showing significant room for improvement in terms of meeting ESC goals for CV prevention. 17% of participants were current smokers at the time of the interview; of those, 42% stated that they intended to quit within the next six months, but only 11% had been referred to a smoking cessation clinic and only 3% had received medication. 44% of respondents were obese and 64% were centrally obese. Of those, 45% indicated that they were actively trying to lose weight, and 52% stated that they were “seriously considering” doing so. In terms of physical activity, only 32% of participants reported getting the recommended 30 minutes of exercise five days a week. While these results are averages, there was wide variation between countries in these and all parameters assessed in the survey, perhaps limiting the broad inferences that can be drawn. Dr. De Backer’s predictable conclusion from these results was that healthy lifestyles have been insufficiently adopted and that more attention is needed, including by developing professional multidisciplinary programs supported by policymakers and health insurers. The results for management of CV risk factors (blood pressure and lipids) were also subpar. Only 54% of patients not on antihypertensive medication had blood pressure below the target of 140/90 mmHg (140/80 mmHg for those with diabetes) and only 43% of those on medication were at goal. The results for LDL cholesterol were even more grim: 11% of patients not on medication had an LDL <2.5 mmol/l (~97 mg/dl) and only 33% were at goal on medications. Dr. De Backer noted that at least based on patient reports, low adherence was not responsible for these findings. He characterized this as a call for action for improved detection, awareness (as many patients with uncontrolled risk factors reported being unaware of this), and management of risk factors.

Questions and Answers

Q: The patient population in this survey was people in primary care who have been diagnosed to be at high risk. The GP has judged them to be at sufficient risk to commit them to lifelong treatment with blood pressure-lowering medications or lipid-lowering medications or they have diabetes. If they’re committed to lifelong treatment for a risk factor, why do so few reach targets?

A: From the clinician’s point of view, they will always say it’s a problem of compliance. If you believe the participants, that’s not the main problem. If you ask patients or others, they’ll call it inertia. Clinicians might be starting prescription of a drug but there’s no follow-up or up-titration if needed. It’s both at the level of the patient and clinicians, and maybe in some countries at the level of health insurance systems, which don’t promote prevention but restrict treatment in terms of cost. Socioeconomic factors play a role. It’s not a simple answer.

Q: Is there data on the dose of drugs patients were prescribed? Were they on an adequate statin dose?

A: That’s a very important question. We have the data but have not yet analyzed it. Now we’re asking the exact same question with the hospital arm and have data on the dosages and kind of statins and other drugs. We hope to present that in the coming months.

Q: I’m not really surprised because real life is always stranger than fiction. You find people taking antihypertensive medication who are not aware of the fact they have hypertension…When you ask patients what normal values or targets are for blood pressure, it’s divided according to age. Most patients over 60 considered 160 or 170 mmHg reasonable. It’s always a problem of patient education…

A: I agree, and these are not representative samples of the high-risk population…I would think these results are better than those in real practice. These participants were more motivated, so I’m not sure in real life the results would be better. It’s not only a question of patient education but clinician education because some are still using levels like 160 mmHg in a 60 or 70 year old. We have to stick to the evidence.

Q: I’m puzzled by your generic conclusion that we need to do better. The question is how. There’s enormous variation between countries. If you look at the first presentation, the country with the most patients identified based on blood pressure was Russia, and Russia had the best blood pressure control. The same goes for Spain with lipids. Why not use Spain to tell Russia how to treat lipids and vice versa for blood pressure?

A: The differences between centers are puzzling. I’m always annoyed by showing the differences with the names of countries because these are not representative centers for whatever country. These are the centers who are willing to join. The differences are real, and we should look and see why they’re doing so well. But I wouldn’t label Russia or Spain as an example. I would look at the specific centers and learn from them.

Identifying and Managing Dysglycemia and Diabetes

Lars Rydén, MD, PhD (Karolinska Institute, Stockholm, Sweden)

Dr. Lars Rydén presented results for the glycemic parameters measured for the first time in this iteration of EUROASPIRE – we were glad to see this addition and believe it is part of a trend toward a more holistic approach to management of cardiometabolic risk. 28% of survey participants had self-reported diabetes, the majority of whom were on oral medications and 41% of whom had an A1c ≤6.5% (58% had an A1c <7%). Among patients who did not report having diabetes, 20% were newly diagnosed based on fasting plasma glucose and/or A1c. In terms of practical implications for screening, Dr. Rydén noted that adding an oral glucose tolerance test (OGTT) to FPG and A1c screening did not substantially increase the “yield” of newly diagnosed patients but that it was much more useful in the 25% of patients that had a high score on the FINDRISC questionnaire. He therefore recommended universal use of FINDRISC as a simple and inexpensive tool to identify people at high risk of diabetes who could benefit from further testing.

Questions and Answers

Q: The risk that one runs in trying to optimize a situation that’s far from optimal is setting the bar too high. I was puzzled by the graph showing control below 6. There’s no evidence to show patients benefit from pushing glucose down so low.

A: It’s a matter of debate. Some of the most convincing evidence comes from the Da Qing Chinese study, which followed IGT patients for 20 years. Those with strict lifestyle control developed diabetes less often and had substantially lower frequency of retinopathy and lower mortality. That indicates that normalized glucose, which they had, is beneficial. There are no drug trials that can show the same results but for type 1 diabetes we’re sure strict control decreases CV morbidity and mortality and microvascular disease. The problem is that so far most studies of glucose lowering drugs have been done in people with long standing diabetes with already established microvascular damage, and it may be more difficult to counteract progress of established injury than to prevent it from appearing.

Q: The precipitating factor for CVD, central obesity, hypertension, etc. seems to be hyperinsulinemia and insulin resistance. Perhaps the only real intervention is one that reduces glucose and insulin like lifestyle intervention and carbohydrate restriction.

A: That remark is music to my ears. I never said you had to use drugs. I referred to the Da Qing study, and there are other indications from Finland and Sweden that it’s possible to intervene with lifestyle adaptation. Giving drugs to lower glucose as a first measure is sort of a failure on our part in not being willing or indeed able to correct lifestyle. We know if you decrease body weight and increase physical activity, the propensity to develop diabetes decreases substantially. We know from comparisons of the impact of lifestyle and metformin in patients with impaired glucose tolerance that lifestyle is considerably more effective because it corrects something incorrect in the body.

Implications for Clinical Practice

Carlos Brotons, MD, PhD (Institute Sant Pau, Barcelona, Spain)

Dr. Carlos Brotons, the only primary care physician among the speakers, summed up his takeaways from the results as “diagnose, awareness, goals.” He argued that the lesson from the fairly high rates of underdiagnosis is that it is important to approach CV risk globally and screen for additional risk factors once one has been identified (e.g., a patient with diabetes and dyslipidemia should be screened for hypertension). He also emphasized the importance of patient awareness and education, as patients are more likely to take responsibility if they are actively engaged and aware of the relevant risk factors and targets. He stressed that this individual approach must also be complemented by a population-based approach that includes government mandates, media campaigns, economic incentives, and school/work settings – this was music to our ears! Dr. Brotons identified adherence and therapeutic inertia as two key factors contributing to the low percentage of patients meeting goals. He referred to results from the West of Scotland Coronary Prevention Study (WOSCOPS) demonstrating the impact of adherence on CV events and mortality and described a study that identified almost twice as many non-adherent patients with electronic monitoring as with a questionnaire, suggesting that measures of adherence in many studies are likely underestimates. In a familiar refrain to those of us in the diabetes world, he described therapeutic inertia as a significant problem that arises from a variety of factors, including overestimation of care, lack of training, clinical uncertainty, and competing demands when caring for patients with many comorbidities. That’s a lot to improve upon, but we thought that raising these challenges was very welcome.

Questions and Answers

Q: The principle of total risk estimation was first put forward in the New Zealand guidelines in 1993 and in the European guidelines in 1994. In Spain you showed that these guidelines are known and yet the majority of GPs don’t estimate patients’ total CV risk. Why is that?

A: About 60% of GPs are not estimating it, perhaps for different reasons. One reason could be limited time in consultations. Sometimes there’s uncertainty because there are several charts in Spain to use, which could create confusion about what is the best chart to use. And some GPs are still skeptical of using CV risk even though we have promoted it in a lot of conferences. We still have to live with skeptical physicians.

Q: I found the findings for hypertension adherence intriguing, where electronic monitoring showed less compliance than was otherwise estimated. Should we not be a bit skeptical of those data from EUROASPIRE that were taken from the interview? If you look at the 42% of smokers who intended to quit and 44% that wanted to start weight loss, would you agree that’s perhaps a bit over-optimistic?

A: When you ask patients, they say they do comply or are going to lose weight or quit smoking. These data are a little overestimated. What we see in real practice is not what is shown here in terms of what patients say they will do. I think there is an overestimation.

Gut Microbiota: New Players in Cardiovascular Medicine

Gut Microbiota in Health and Disease

Fredrik Karlsson, PhD (Örebro University, Örebro, Sweden)

Dr. Fredrik Karlsson reviewed the state of the science with regard to the hot topic of the gut microbiome and its potential impact on disease. He explained that metagenomic techniques allow researchers to sequence the genomes of bacteria found in stool samples and compare to reference genomes to determine species abundance, gene abundance, and associations with various disease states. He discussed his lab’s work constructing a gene catalog of 11 million bacterial genes and analyzing the similarities and differences between cohorts from different regions (European cohorts had more total genes than US and Chinese cohorts, almost 300,000 genes were shared by half the population, and almost none were shared by everyone). He reviewed evidence suggesting that reduced microbial diversity is associated with metabolic disease, and at least in animal studies, specific classes of microbiota have been shown to play a causal role in obesity and metabolic disease. He shared several examples of studies demonstrating varying microbiota composition in patients with atherosclerosis, diabetes/impaired glucose tolerance, and obese patients post-bariatric surgery compared to controls. Of course, all of the human research remains at the level of broad associations, and as Dr. Robert Ratner (ADA, Alexandria, VA) laid out in an informative talk on this subject earlier this year, much work needs to be done before specific connections can be made between particular bugs and human disease.

Questions and Answers

Q: I didn’t follow your last point. Were you trying to make the point that unless you had those bacteria you wouldn’t gain weight or were those an undesirable consequence?

A: The increase in proteobacteria [following bariatric surgery] has not previously been associated with good health. We don’t exactly know how to interpret it. When we introduce them into germ-free mice, they don’t gain as much weight as with obese flora. It’s an unsolved puzzle whether it’s a good change or detrimental.

Q: Have you done research into the effects of antibiotics?

A: We have not, but others have and see large changes. They’re typically transient, so it goes back to the original composition. I’m not too familiar with the exact evidence, but the overuse of antibiotics and the association with metabolic disease is something that’s interesting.

Q: Gut microbiota are associated with a number of diseases. Is it a different composition in different diseases or is it an inflammatory change that’s the same in all diseases?

A: I would say it’s a young field so we haven’t done those kinds of comparisons. It’s also difficult to compare between studies. Rich, diverse microbiota is associated with low inflammation and a healthy state. Specific diseases might have a bloom of specific microbes.

Q: What effect does public water treatment with chlorine have?

A: I have never seen a study on that. My gut (chuckles) feeling is it’s a very low concentration of chlorine that you would have in the gut compared to when it’s in the treatment plant.

Q: Are there clinical consequences from altering the microbiome using prebiotics?

A: Fibers generally increase diversity. A diet rich in vegetables and fiber results in diverse microbiota and is generally good for health. I can’t say one specific prebiotic has come out as particularly good.

Personalized Nutrition: Has the Time for Action Come?

José Lopéz Miranda, MD, PhD (University of Cordoba, Spain)

Dr. José Lopéz Miranda delivered an intriguing presentation on the possibilities of personalized nutritional recommendations based on specific genetic variants. He explained that data suggests that it is an interaction of genetic variation and dietary patterns that ultimately determines cardiovascular risk. For example, in the PREDIMED study, people with a particular genetic variant had a significantly decreased incidence of stroke on a Mediterranean diet. In another study, a certain group of patients benefited more from a high monounsaturated fatty acid diet, while another group did best with a high complex carbohydrate diet in terms of LDL cholesterol. The same holds true for metabolic disease: Dr. Lopéz Miranda presented evidence that certain SNPs are associated with insulin resistance when they interact with saturated fat, which means that genetic variant only acts as a risk factor in the context of a high-fat diet. While he cautioned that such findings need to be replicated and the mechanisms better understood, the vision of a future where patients could be given specialized dietary recommendations tailored to their genotype is a very promising one given the current level of uncertainty in nutrition science (we will have a very interesting interview with Dr. Anne Peters coming out related to this question – let us know if you’d like that sent to you directly).

Questions and Answers

Q: I agreed with your point about personalized nutrition but I can’t say I agree with you on the best diet for metabolic syndrome. We’ve seen numerous studies that have put people on a low-carb diet that’s higher in saturated fat and seen improvements. Diabetes has actually gone into remission. From what I’ve seen, to think we’re still as the default basing our dietary paradigm on a low-fat diet that’s relatively high-carb will surely just exacerbate the situation. It just seems irresponsible and bordering on malpractice.

A: I will show you just the evidence we have, many of which comes from large dietary intervention studies. It has shown that when you adopt a Mediterranean diet you reduce the risk of diabetes. When you adopt a vegetarian diet you also reduce risk. One of the first studies I showed is just to increase physical exercise, reduce saturated fat, and increase the content of complex carbohydrates. That was based on the largest dietary intervention study with more than four to five years of follow-up.

Q: I’m trying to connect what you said to the gut microbiota? How is that affected?

A: We know that when you change specific nutrients in the diet, you could change the composition of the gut microbiota. There’s some evidence, but the next question is are these changes induced by diet associated with specific changes in components that have an effect on metabolic syndrome risk, and we don’t know.

Q: We have a couple of different breeds of livestock and genetically modified grains. How does that factor in? If I’m eating something that’s resistant to certain parasites and it affects my microflora, I’m reducing my fat intake but I might be introducing something else detrimental. Now we’re seeing lots of genetically modified products.

A: I’m not an expert in this area but it could be a possibility.

Is Obesity Always Bad?

Obesity is Always Bad in Cardiovascular Diseases

Dan Gaita, MD, PhD (Institutul de Boli Cardiovasculare, Timisoara, Romania) and Wolfram Doehner, MD, PhD (Charité Universitätsmedizin Berlin, Berlin, Germany)

In this provocatively titled debate, Dr. Dan Gaita (pro) and Dr. Wolfram Doehner (con) agreed on the key points: (i) obesity constitutes a major risk factor from a public health perspective and (ii) there is work to be done to better define and measure it. Dr. Gaita maintained that broad focus during his initial presentation, presenting the sobering statistics on rising obesity prevalence and its association with cardiovascular risk and mortality (“what does bad mean? Bad means death”). Dr. Doehner aimed to add a bit more nuance to the story by distinguishing between primary prevention in the healthy population and secondary prevention in patients with existing CVD, in which case there is substantial evidence that higher BMI is associated with lower mortality. In his rebuttal, Dr. Gaita argued that this “obesity paradox” is likely an example of the limitations of the current BMI-centric approach. He also noted that such studies only constitute a snapshot in time and do not account for the long-term risks of prolonged obesity. Dr. Doehner granted that BMI carries limitations but argued that given its current status as the main criterion for obesity, studies that rely on it should still be taken seriously. He also noted that voluntary weight loss appears to be just as risky as involuntary weight loss in these high-risk patients and highlighted the findings from the Look AHEAD trial showing no reduction in mortality despite improvement in a number of weight-related parameters. In our view, this debate illustrates the limitations of our current knowledge of obesity and perhaps provides some nuance in how cardiologists should approach caring for high-risk patients. However, from a public health perspective, we hope that such debates will not detract from the larger call for societal solutions to address the undeniable negative consequences of the obesity epidemic.

Questions and Answers

Q: Are there any studies on obese patients where the medical community has used aggressive pharmacological interventions to bring down LDL 50%, prevent diabetes, reduce albuminuria, give an ACE/ARB, etc.? If you bring down the risk of obese patients with pharmacological interventions, does that bring their risk down?

Dr. Doehner: If you can improve their metabolic profile you will improve outcomes. I can’t give you studies off the top of my head. All I’m arguing is that weight loss is usually achieved by starvation, which means caloric restriction and catabolic aggravation.

Q: I don’t want them to lose weight. I’m talking about intervening earlier to bring down cholesterol and other factors and whether that puts them in a less risky position even than the non-obese population.

Dr. Doehner: It’s difficult to give a clear answer. In heart failure, lowering cholesterol is not good either.

Q: We have studied obese patients without comorbidities and showed that obesity even without comorbidities is a disease. You have to look beyond obesity and look at epicardial fat, etc. Maybe the subgroup without that is different. Does the group you studied have no comorbidities?

Dr. Doehner: I’m not saying obesity is something not to worry about. When obese patients have no other clearly established CVD already, that’s the case for primary prevention. It’s better if people are prevented from becoming obese. If those patients lose weight, whether they improve back to the mortality risk of those with a lower BMI is less clear. As seen from Look AHEAD, losing weight didn’t improve mortality.

Q: Some obese patients may already be there based on a CT scan and CAC score even though they are asymptomatic.

Are There Metabolically Healthy Obese Patients?

Nicolae Hancu, MD (Iuliu Hatieganu University of Medicine and Pharmacy, Cluj-Napoca, Romania), Eberhard Standl, MD (Munich Diabetes Research Institute, Neuerberg, Germany), and Laurence Sperling, MD (Emory University, Atlanta, GA)

In this somewhat more pointed debate relative to the first in the session, Dr. Nicolae Hancu and Dr. Eberhard Standl argued that the metabolically healthy obese constitute a real and important subset of the obese population, while Dr. Laurence Sperling argued that the concept is “either mythical or more likely a misinterpretation.” Dr. Hancu acknowledged that the definition and prevalence of metabolically healthy obesity remains unresolved, as does its predictive power for cardiovascular risk, diabetes, and mortality. However, he and Dr. Standl explained that there are clear differences between abdominal and lower-body fat (likely due to differential gene expression) that have important metabolic consequences. He presented results from mechanistic studies demonstrating very different responses from adipocytes in the two locations during weight gain – essentially, gluteal adipose tissue acts as a “metabolic sink” that protects against excessive free fatty acid exposure and insulin resistance. Dr. Hancu also noted that fitness appears to play a similar role in “training” adipose tissue to behave in a more functional way. While Dr. Sperling conceded that this may be an interesting academic debate, he argued forcefully against the concept of healthy obesity as it pertains to clinical practice. He noted that most of the so-called metabolically healthy obese do have one or two cardiovascular risk factors and suggested that many of the rest likely have sub-clinical disease. He, like Dr. Gaita in the first debate, also emphasized the importance of a long-term perspective; as he put it, obesity is like a ticking time bomb that can be silent for decades but will eventually lead to damage. Even if some obese patients are truly metabolically healthy, Dr. Sperling also reminded the audience that metabolism is not the whole story, and that obesity exerts effects on every organ system and on society as a whole due to the resultant excess healthcare costs.

Obesity and Heart Disease

Current Understanding of Visceral Obesity

Jean-Pierre Després, PhD (Laval University, Quebec, Canada)

Dr. Jean-Pierre Després argued that “body shape matters a lot more than body size” when considering the risks associated with obesity, outlining the “minestrone soup” of abnormalities related to excessive visceral adipose tissue (VAT). Like Dr. Nicolae Hancu and Dr. Eberhard Standl the previous day, he explained that there are clear functional differences between VAT and lower-body subcutaneous fat: while the latter can act like a metabolic sink and expand to protect lean tissue from excessive lipid accumulation during weight gain, the former becomes hypertrophic and inflamed, leading to lipid spillover into the heart, liver, skeletal muscle, etc. Not surprisingly, VAT is significantly associated with excess liver fat, diabetes, and cardiovascular disease. Given these associations, Dr. Després cautioned that the evidence showing a recent plateau in BMI in the US may be misleading, as waist circumference has continued to increase. Overall, he advocated a more comprehensive approach to obesity diagnosis and treatment that takes into account waist circumference, cardiorespiratory fitness, nutritional quality, and physical activity habits.

Questions and Answers

Q: Nutrition and physical activity are important variables to tackle to reduce visceral fat deposition. What about the effect of sleep? That’s commonly neglected, and we’re getting less in terms of quantity and quality.

A: Yes, we’ve been very interested in CVD and diabetes, but there are additional health outcomes associated with excess visceral fat like brain health and respiratory disease. Visceral obesity is associated with fat in the tongue, which is an interesting ectopic fat depot that causes a prejudice to breathe in, and sleep apnea is the major consequence. Then the patient gets into a vicious cycle where they can’t sleep, they’re more sedentary, and that exacerbates the metabolic profile.

Q: You didn’t emphasize the most important issue of epicardial fat.

A: I’m not sure epicardial fat is the main problem. There’s evidence indicating a robust relationship between that and clinical outcomes, I fully agree. To say it’s the sole player in comorbidities related to ectopic fat, I’m not sure. Type 2 diabetes risk is probably related to dysfunctional subcutaneous and liver fat. Liver fat plays a key role because a fatty liver pumps out more glucose and triglycerides. If you want to raise that point, it is important, but it’s not the sole player.

Q: We have done studies where epicardial fat correlated with coronary artery calcium on a CT scan.

A: I agree, if that is your favorite ectopic fat depot I believe you. I think we should pay attention to all ectopic fat depots, and depending on the clinical manifestation, their relative weight may differ. Time will tell if I’m right.

Metabolically Healthy Obesity From Korean Cohort Study: False or True?

Seong Hwan Kim, MD (Korea University, Ansan, South Korea)

Like Dr. Laurence Sperling in the previous day’s session, Dr. Seong Hwan Kim argued that metabolically healthy obesity (MHO) is “simply a myth” and is more akin to a transient state of lower risk before conversion to traditional unhealthy obesity. He presented evidence demonstrating initially no significant difference in cardiovascular risk between MHO participants and normal weight controls, but an increase in risk beginning after a decade that was maintained up to 30 years of follow-up. He also presented data from the Korean Genome and Epidemiology Study (Ansan cohort) showing that while there was no significant difference in carotid intima media thickness in MHO vs. normal weight individuals, there were differences in other cardiac parameters. In addition, while there was no difference in hypertension rates between the groups initially, risk began to increase after four years. Interestingly, Dr. Kim noted that when adjusted for fitness, the increased risk with MHO disappeared, suggesting that at a given level of fitness, it may be true that MHO is benign. Overall, Dr. Kim emphasized that obesity is a clear risk factor for CVD at the population level and that while the existence of an MHO phenotype demonstrates the heterogeneity of obesity, current evidence suggests that it is not completely innocuous.

PCSK9 Inhibitors and Lipid Management

PCSK9 Inhibition in Lipid Management: Final Steps Toward Implementation into Clinical Practice (Supported by Physicians’ Academy for Cardiovascular Education)

Michel Farnier, MD, PhD (Point Medical, Dijon, France), John Kastelein, MD, PhD (University of Amsterdam, Netherlands), Paul Ridker, MD (Brigham & Women’s Hospital, Boston, MA), and Kausik Ray, MD (St. George’s University of London, UK)

To an absolutely packed room, Dr. Michel Farnier introduced this session by highlighting the impressively accelerated research timeline for PCSK9 inhibitors, from the discovery of the PCSK9 gene in 2003 to the approval of the first two medications this year. Dr. John Kastelein then provided a clear and concise overview of the drugs’ mechanism of action and the discovery of genetic mutations illustrating the dramatic effect of PCSK9 on LDL cholesterol and long-term cardiovascular risk. In his presentation, Dr. Paul Ridker focused on the ongoing cardiovascular outcomes trials for PCSK9 inhibitors, stressing to the audience that “it is incumbent on us” to ensure patients remain enrolled in the trials even when the drugs are on the market – this potential disruption to trial integrity was one of the concerns about an early approval raised during the FDA Advisory Committee meetings for Praluent and Repatha. He also suggested that payers will likely want to see data from these trials on event reduction and long-term safety before they “move the way we want them to.” Dr. Ridker also noted that the SPIRE CVOT for Pfizer’s bococizumab includes a broader population of patients with diabetes (those without established coronary artery disease) compared to the other two trials, suggesting that they should provide a somewhat better indication of the class’ profile in the general type 2 diabetes population. Dr. Kausik Ray closed the session by offering his vision of the expected advances in lipid management over the next five years, including a “revolution” in the management of familial hypercholesterolemia (FH), a move toward earlier screening and treatment of high LDL, and more sophisticated risk algorithms to determine which patients require additional LDL-lowering therapy beyond statins. In an echo of the ongoing debates in the type 2 diabetes field, he suggested that it would be more cost-effective to initiate treatment earlier in the course of disease and that the current approach of treating disease rather than preserving health is sub-optimal.

Questions and Answers

Q: You showed a new cut of the JUPITER data that puzzled me. Some of the patients had LDL increase while on a statin still had a benefit in outcomes vs. those on placebo. They may have been non-compliant, but what’s going on?

Dr. Ridker: Pure trialists might call this the healthy cohort effect. This is six-month or one-year LDL, so maybe they were partially compliant. We’ve seen that for other parameters. The other side of the graph is what I’m concerned about. If you have an 80-85% reduction with a statin, I’m not sure how much more benefit another drug is going to give. If you’re compliant with a statin and only getting a 25% reduction, payers might want to go with a PCSK9 inhibitor.

Q: You mentioned the anti-inflammatory effect of statins. PCSK9 inhibitors don’t have that. What’s the mechanism of risk reduction then?

Dr. Ridker: I’d argue that lowering LDL is a phenomenally good way of lowering risk. I’m not surprised at all. My point was that the beauty of these parallel experiments is that PCSK9 inhibitors don’t lower CRP but they’re powerful drugs that mimic part of what statins do, so it’s a good test of the LDL hypothesis.

Q: PCSK9 and the LDL receptor are designed to regulate intracellular cholesterol, not fight atherosclerosis. What is happening with the excess intracellular cholesterol that the hepatic cell is dealing with when PCSK9 is inhibited?

Dr. Kastelein: Many mechanisms regulate the intracellular pool of cholesterol, such as increased bile acid synthesis and excretion of cholesterol into bile, for example. That’s one of many mechanisms by which the body protects itself from high cholesterol. The pool is so well regulated in the hepatocyte, I’m not aware of any toxicity associated with that. It’s also never in the form of cholesterol but all in the form of cholesterol esters in the hepatocyte. Cholesterol esters are bad when they’re in the arterial wall, not in the hepatocyte.

LDL-C Management: Can We Redefine What Is Possible? (Sponsored by Sanofi/Regeneron)

John Kastelein, MD, PhD (University of Amsterdam, Netherlands) Kausik Ray, MD (St. George’s University of London, UK) and Robert Eckel, MD (University of Colorado, Aurora, CO)

To another packed room, Dr. John Kastelein framed the discussion by reviewing the significant remaining unmet need for more LDL-lowering therapies and the impressive body of evidence (“the most evidence-based regression line we have”) demonstrating a direct relationship between LDL reduction and improved outcomes. Both Dr. Kastelein and Dr. Kausik Ray offered a slightly sarcastic critique of the latest ACC/AHA guidelines that did away with specific LDL targets, and Dr. Ray’s presentation focused on the evidence supporting a treat-to-target approach on an individual and a population level. Dr. Kastelein then returned to the stage to discuss how the results of the IMPROVE-IT trial (evaluating statins + ezetimibe) demonstrated that the LDL/outcomes regression line is applicable beyond statin therapy and that lowering LDL to levels below the current targets appears to be beneficial. The benefits and potential risks of very low LDL was one of the main concerns raised both at this conference and at the AdComm meetings for the PCSK9 inhibitors and should be one of the most important questions answered by the ongoing outcomes trials. Dr. Robert Eckel closed the session by outlining the three groups of patients that he believes represent the remaining unmet need for LDL-lowering therapies: FH patients, statin-intolerant patients, and a broader high-CV risk population (including plenty of patients with type 2 diabetes). He reviewed the clinical evidence in support of PCSK9 inhibitors in each of these populations and reminded the audience of the importance of successfully completing the ongoing trials. He also offered somewhat of a defense of the American guidelines, saying that they represent the most evidence-based approach based on current data, but acknowledged that he too uses goal-setting behavior in practice. 

Questions and Answers

Q: LDL is probably the most important risk factor but others may play a role if LDL is lower.

Dr. Ray: Right, we talk about global risk, and one thing that would determine whether you think about further treatment is if HDL is low.

Q: Instead of confrontation, why not work on the global guidelines?

Dr. Eckel: I can’t speak for what’s going on at the ACC and AHA. The IMPROVE-IT and PCSK9 trials will demand that certain LDL levels be achieved. I support the evidence that was used. I believe in goal-setting behavior in the real world.

Dr. Cannon: There are a few guidelines that have been joint, and we fortunately have been learning from the ESC guidelines and evolving. Here the evidence has evolved, so it’s an exciting time in the field. We have a solid evidence base saying that getting statins and using intensive dosing is a good step one, then rechecking and adding therapies to titrate to goal is the second phase.

Dr. Kastelein: There are lots of politics and egos involved.

Q: If we look at the level of LDL after treatment in IMPROVE-IT, was it similar to ODYSSEY or even lower?

Dr. Cannon: An interesting distinction is that both ODYSSEY LONG TERM and IMPROVE-IT ended up at around 1.3 mmol/l (50 mg/dl) but IMPROVE-IT treated people at goal. That’s because ezetimibe is pretty modest. All the ODYSSEY trials are people not at goal. PCSK9 inhibitors are so strong, they can take people from way up high to low levels. People with moderate elevations go even lower.

Q: Why was the LDL in the placebo arm of ODYSSEY LONG TERM so high?

Dr. Kastelein: Many have heterozygous FH, so they simply have high LDL despite statin therapy. It’s a consequence of the randomization scheme. LDL is defined by the entry criteria, so by definition it’s different in every trial. I believe the baseline LDL in the outcomes trials was much lower?

Dr. Cannon: I don’t know, but there you have people with high risk treated according to current guidelines, so I think it’s more like 75-80 mg/dl.

Dr. Eckel: It’s above 70 mg/dl for three of them and above 100 mg/dl for the other.

Q: With the alirocumab data, what about HDL?

Dr. Eckel: Triglycerides come down minimally and HDL is up in the 5-10% range. You clearly wouldn’t select a PCSK9 inhibitor to raise HDL or lower triglycerides.

Q: You mentioned that lack of PCSK9 is associated with fatty liver. As far as I know it’s the opposite?

Dr. Eckel: The specific genetic disorder for low LDL is associated with fatty liver.

Q: Was there any suggestion that alirocumab lost efficacy over time?

Dr. Cannon: There was no loss of efficacy. One of the big surprises in the initial trials was that by the end of the study 85-90% of patients were taking the injections a year later. That’s better than in many trials of pills. It’s an interesting finding to say this new mode of lipid therapy is tolerated and patients are compliant. And those are settings where patients are blinded to lipids. In open-label extensions where patients see their lipids, 98% are compliant. Any changes on the graph are some people falling off.

Q: Why were there two different LDL targets in the same study?

Dr. Ray: For different levels of risk.

Q: Can you reiterate what you said about neurocognitive side effects with low LDL?

Dr. Eckel: With very low LDL, defined as less than 25 mg/dl, there was not a hint that that level was related to any neurocognitive effects.

Q: Which PCSK9 inhibitor would you give your grandmother?

Dr. Cannon: The key is what is the LDL? It’s about LDL and the CV risk equation. The data is dramatic across the whole class. It’s using the basics of statins and ezetimibe if that’s enough and then you can use any agents. The problem is real inertia in even adopting statins. Hopefully all these new data will invigorate us to look at LDL, dig in, and revisit everyone’s risk.

Dr. Eckel: If you’ve got a patient who wants to inject every four weeks, evolocumab is your choice. If you want treat-to-target therapy, the option for 75 mg vs. 150 mg dosing [with alirocumab] is important.

Q: How important is the small dense LDL story?

Dr. Eckel: Measuring the size of LDL has no value in clinical decision-making.

Dr. Cannon: The number is key. There are selected patients where advanced lipid testing can be helpful.

Q: Does ezetimibe affect PCSK9?

Dr. Cannon: Yes, as LDL goes down PCSK9 goes up.

Q: At an age above 75 or 80, does LDL reduction above 50% make a difference?

Dr. Cannon: That’s where it’s most impressive. We saw an 8.5% absolute reduction in events in people over 75 by adding ezetimibe. The notion of backing off because of potential side effects is really opposite. These are high-risk patients who benefit greatly.

Dr. Ray: We wrote the PROVE-IT analysis, and if you were over 70 and had an LDL less than 70 mg/dl, the gain was an 8%-9% absolute risk reduction. Your LDL needs to be lower than your age.

Q: What is the ideal starting dose of alirocumab?

Dr. Eckel: If you look at the trials, the entrance criteria was above 70 mg/dl for three out of four. You can think of three pillars: FH above 70 mg/dl, very high-risk, and statin-intolerant. Look at the type of patient and make an educated decision. We’re inadequately informed about how low we should go.

Dr. Cannon: If LDL is above 130-140 mg/dl, then use the higher dose. In the middle range, you could use the lower dose.

PCSK9 Inhibitors and Evidence-Based Outcomes: What Does the Future Hold? (Sponsored by Pfizer)

Christie Ballantyne, MD (Baylor College of Medicine, Houston, TX), Gilles Lambert, PhD (University of Nantes, France), and Jean-Claude Tardif, MD (University of Montreal, Quebec, Canada)

Dr. Christie Ballantyne opened by reviewing the enormous progress in the lipid field over the past few decades (contrasting it to when “people thought I was crazy” for focusing on lipids) and emphasizing the significant remaining residual risk. He notably contrasted the advances in lipids with obesity, which has gone in the wrong direction, and diabetes, where the lack of a clear connection between glucose lowering and event reduction has been disappointing. Dr. Gilles Lambert, who conducted several seminal research projects in the PCSK9 field, then delved deeply into PCSK9 inhibitors’ mechanism of action and the benefits of human monoclonal antibodies in general (high specificity and potency, low dosing frequency, low immunogenicity). Dr. Jean-Claude Tardif focused on the clinical data for the class and the remaining questions to be answered in the ongoing outcomes trials. He referred to the “exciting” post-hoc analyses suggesting hope for large event reductions, though he cautioned that the post-hoc nature of the analyses and the low number of events are significant caveats and that the results should simply “whet our appetites” for the definitive outcomes trial results.

Questions and Answers

Q: Is there danger in getting LDL too low?

Dr. Ballantyne: We haven’t seen that so far in the trials. The statement in the guidelines about avoiding LDL below 40 mg/dl is not evidence-based. If you look at the data from JUPITER and IMPROVE-IT, many people got LDL below that and they had the best outcomes. We have never gone into the single digits or under 20 mg/dl, so that’s one thing that’s exciting about the outcomes trials. The ODYSSEY program starts with a lower dose, but the evolocumab program starts with the full dose and says don’t worry about LDL. The bococizumab program starts with the full dose and says you can down-titrate. Because of that we’ll actually get answers. So far we haven’t seen anything.

Dr. Tardif: I agree, which is why I mentioned the genetics. We have patients with loss of function mutations who have LDL values of 15-20 mg/dl from birth, and they have normal renal function, they have jobs and kids.

Dr. Ballantyne: And they don’t get heart disease.

Dr. Tardig: There have been signals that are probably nothing about cognitive dysfunction. That’s why we need to complete the trials to get robust signals on efficacy and full safety.

Q: Statins increase PCSK9, so that will oppose their effect. Is that the reason why there’s such a small additional benefit as you increase the statin dose?

Dr. Lambert: That’s certainly the main reason why when you increase the dose, the percent reduction with the first dose is big and then it reduces. It’s a curve. PCSk9 inhibitors alone or on top of statins probably stop that.

Q: If we have a class that really lowers LDL, SPIRE-2 is the trial to test that since they’re coming in with higher levels.

Dr. Tardif: In the trio, that’s the only trial focusing on that group. I think it’s very important because a lot of patients, for all sorts of reasons, don’t reach targets despite treatment, so we need to do something. What we’re seeing now is that we have two agents approved but we realize it’s not sufficient to get approved by regulators. You need payers and formularies to agree. We’re seeing all sorts of algorithms to slow the process down. I think people will want them to be used where you get more bang for your buck. Patients with an LDL above 100 mg/dl is one population where we need new agents.

Q: What is the experience with these agents in terms of safety?

Dr. Ballantyne: They look very safe. There’s the occasional injection site reaction, which is minor. We mentioned the questions about a signal for a neurocognitive effect, but the numbers are extremely small. There’s not any clear signal but we don’t have data to rule it out. It’s quite nice when dealing with therapies when the question is that we don’t have enough data to be certain if there’s a side effect of not. With most drugs, you can figure out the side effects after 500 patients. It’s important if we will use these drugs in thousands of people or perhaps millions that we figure out if there are infrequent side effects, and we can only evaluate that with large exposure. The outcomes trials are equally important for safety. We found out late that there could be a slight risk of diabetes with statins. The power of three programs is a wonderful thing.

Q: I believe I noticed one study done on background statins and ezetimibe. Were the others just statins?

Dr. Tardif: The cornerstone of therapy remains statins, but they are allowed to use ezetimibe if the clinician decides to in all the studies.

Q: Is there evidence of immunogenicity?

Dr. Lambert: It hasn’t been a problem. You can count the reported cases on one hand.

Dr. Tardif: [To sum up,] there’s lots of potential here, including in patients with high risk because of diabetes or other risk factors. To confirm our hopes, these trials need to be completed. They need to be fully recruited and patients retained to the final stages. The story is not ended.

PCSK9 Inhibition: The Treatment (R)evolution for Dyslipidemia in High-Risk Patients (Sponsored by Amgen)

John Chapman, MD, PhD (Hôpital de la Pitié-Salpêtrière, Paris, France) and Kausik Ray, MD (St. George’s University of London, UK)

To another completely full room, Dr. John Chapman reviewed the immense importance of cholesterol to the body, the demonstrated causal role of LDL in atherosclerosis, and the genetic studies that led to the identification of PCSK9 as a promising therapeutic target. Dr. Kausik Ray then discussed the three main areas of clinical unmet need (FH, statin intolerance, and high LDL and CV risk despite maximally tolerated statins) and the need to focus on global risk when assessing the need for new therapies. As in his earlier presentation, he emphasized the importance of addressing elevated LDL earlier on, when it is possible to shift the life expectancy curve much further. He also highlighted the very positive safety profile of PCSK9 inhibitors in clinical trials thus far, suggesting that the ongoing outcomes trials should provide even clearer reassurance. In terms of event reduction, he cautioned that we do not yet know the answer but described the ~50% reduction seen in the phase 3 post-hoc analyses as very encouraging.

Questions and Answers

Q: What is the role of PCSK9 outside the liver?

Dr. Chapman: It relates to four tissues: the liver, intestine, kidney, and brain. The body has very fine control of intracellular cholesterol. At levels below five moles/cell, SREBP is activated and LDL receptor levels increase. There’s very fine control, and that is true in the intestine and kidney. We know there’s an exception in the liver because this is the organ with the highest cholesterol flux. It’s the organ that functions to excrete cholesterol. There’s a role for PCSK9 in modulating LDL receptor expression in the intestine and kidney but the level of the receptor is substantially less. In the brain, there appears to be a potential role during embryogenesis. In adults, the size of the monoclonal antibody is more than 10 times larger than any molecule able to cross the blood-brain barrier. There’s no evidence for traversing of that barrier.

Q: How important is constant suppression of PCSK9 and LDL over time?

Dr. Ray: There’s an argument that if you look at apheresis, you get big fluctuations in LDL. The only data is a paper looking at variability of LDL that found if you have greater variability, it’s associated with worse outcomes. That’s in the context of an oral agent. The most common reason for variability is not taking the drug. That’s the data we have.

PCSK9 Inhibition: An Important Step Forward in Treating Dyslipidemia in High-Risk Patients

Kausik Ray, MD (St. George’s University of London, UK) and Erik Stroes, MD, PhD (University of Amsterdam, Netherlands)

Q: Is lower LDL better?

Dr. Ray: In my opinion, the answer is absolutely yes. It starts with whether lower LDL translates to a lower risk of events. We have observational data and RCT data in statin and statin + ezetimibe treated patients. Many years ago, we did an analysis from PROVE-IT to ask the same question. Those in the lowest quartile had the lowest risk of CV events. That was reproduced in several studies, which also showed that it was safe. In the recent study, we got to the lowest we’ve ever seen in an outcomes study, below 30 mg/dl. There was no relationship between on-treatment LDL and any side effects.

Dr. Stroes: When we first saw the levels coming back so low, we got nervous. Some protocols even said oh no, we have to do something. There’s a disease caused by a genetic defect in apoB where patients have an LDL of 10-15 mg/dl. The patients are healthy and they have fewer CV events and flexible arteries. They have some hepatic issues because of cholesterol accumulation. That kept us going. We look forward to the safety data.

Dr. Ray: That’s true with loss of function of PCSK9 as well. We don’t see any gradation of risk with the gradient of LDL. When side effects appear, it’s at random. It’s exactly the same reassuring trend as in IMPROVE-IT and OSLER.

Dr. Stroes: And evolocumab and alirocumab are completely comparable. What about the fear for 10-year complications?

Dr. Ray: I can’t really comment on the long-term safety with this, but experiments of nature don’t suggest that.

Q: Clinically, we do see patients with neurocognitive decline related to statins. Is it the amount of decline in LDL that may be of consequence?

Dr. Ray: I don’t think we have any randomized trial evidence. We’ve all seen patients complaining of something, but we don’t have a control group. People have diabetes, hypertension, lots of issues that could be risk factors. Unless you have a control group, you can’t blame it on the statin. If someone gets a side effect, it’s the drug. No analysis has ever shown that it’s on-treatment LDL or the percent reduction. As yet there is no safety signal. If one ever appears, and I don’t believe it will, it won’t be LDL. And because monoclonal antibodies are so specific, I doubt it.

Dr. Stroes: It may turn the other way. There are so many new findings that dementia has a 30%-40% attributable risk to CV risk factors. We’re lucky there are neurocognitive sub-studies, so we may be surprised. If they massively lower CV risk, it could go the other way.

Dr. Ray: The study Amgen is doing will be very interesting. We will get objective evidence.

Q: Evolutionarily, there’s an argument that an LDL of 1.2 mmol/l is when the receptors are saturated. We need lipids to solidify membranes, so physiologically that doesn’t make sense.

Dr. Stroes: I love that discussion. We tried to go to zoos and look at animals. Big mammals have an LDL of 0.5 mmol/l. Babies, when their brains are developing, have an LDL of 15-20 mg/dl. There’s only one example in evolution where high LDL was good. Two or three centuries ago when inflammatory disease caused mortality, FH families had a survival benefit. In an era where there were no antibiotics or hospitals, high cholesterol seemed to be protective and acted as a sink for inflammation.

Q: What about the rapidity of decreasing LDL. If you lower blood pressure very fast, it may not be beneficial. Is that of concern?

Dr. Ray: If you take apheresis as an example, it brings LDL down very quickly and we accept that. Look at two randomized trials, PROVE-IT and A-Z. PROVE-IT enrolled patients straight after ACS, and within 30 days had a significant LDL drop to below 70 mg/dl. In a comparable trial, the PIs were worried, so they had a placebo arm for four months, then started 20 mg simvastatin, then escalated. If you look at the early phase, there was absolutely no curve separation. If you look at the more stable phase, it was identical to PROVE-IT. If you pick a high enough risk condition, getting LDL down quickly is a benefit, not a harm.

Q: My question is about neurocognitive function. There seems to be an initial effect with statins and then it wanes.

Dr. Stroes: That point is valid. If you acutely control glucose, for example, some strange side effects occur. From that point of view, you can imagine why lowering so acutely could be bad. We should wait for outcome data. If you look at the CV benefit vs. the potential risk of a temporary side effect, the likelihood of help vs. harm is still very good. Yes, we need long-term data, but you have to consider that acute rapid lowering has a lot of benefit.

Q: In which patients would you use a PCSK9 inhibitor?

Dr. Ray: FH, high CV risk and true statin intolerance, and people with high absolute risk whose LDL remains high despite maximum lipid modification treatment.

Q: Why are the majority of people not treating LDL more aggressively?

Dr. Ray: I’d put that to the audience. Are the drugs too expensive? Are you worried about high doses?

Dr. Stroes: We have something that lowers LDL that appears safe. We all know PCSK9 inhibitors will never in my lifetime be the first choice. The first choice is power statin therapy with ezetimibe, which will be generic in a year and a half. If we ever start to think about PCSK9 inhibitors, we will be forced to have optimized generic therapy. 70% of high-risk patients weren’t using maximum therapy. The second point of the advent of PCSK9 inhibitors is we will be forced to up-titrate current medication.

Dr. Ray: Work on lipids will force people to start thinking about prevention and optimizing current therapies.

Q: Do patients mind taking injections?

[Audience poll: Only a few say yes]

Dr. Stroes: I fully agree. In the beginning I said it would be a challenge, thinking of diabetics who I told had to switch and it was “oh no, not insulin.” When we started the trials, we have seen enormously few adverse effects. Some studies included patient satisfaction measures, and patients love it because they like it better than taking pills every day. You can imagine that a subcutaneous injection with a nice device every two weeks is easier. So I can say no, if you do it correctly you won’t run into lots of fear. Pills will remain the first choice. We haven’t resolved the adherence issue yet.

Dr. Ray: It comes back to patients’ perception of risk. Last year there was lots of data about long-term safety, there were press releases, and some people found my email address and I had lots of patients asking about it. They had FH, multiple stents, and some couldn’t tolerate statins. If you read what they said, your heart goes out to them. They’ve tried multiple medications and they have progression of disease. Who will mind an injection? Those who won’t be taking it in the first place. Those who really understand the risk will be fine.

Q: Have you seen people develop antibodies?

Dr. Ray: At least in the short term it’s a very low percentage. In the long-term studies you have 70,000 people for three years. We’ll know then and know the clinical relevance.

Dr. Stroes: It’s a pivotal question because sometimes you completely lose efficacy. We expect a minute rate with human antibodies. Yes there have been antibodies, even a few neutralizing ones, but the response is not affected. We will have to wait for the huge databases.

New-Onset Diabetes Mellitus in the IMPROVE-IT Trial

Michael Blazing, MD (Duke University, Durham, NC)

Dr. Michael Blazing presented results from an analysis of the IMPROVE-IT trial showing no significant increase in new-onset diabetes with ezetimibe added to simvastatin vs. simvastatin alone. As he noted in his introduction, this analysis arose from the concerns about increased diabetes risk with statins. It would seem to offer a good opportunity to test whether that risk is due to statin-specific effects or to LDL-lowering in general (a theory Dr. John Kastelein referred to in his presentation earlier in the same session). In IMPROVE-IT, there were 1,414 cases of new-onset diabetes (13.3% of the study population), 720 in the ezetimibe/simvastatin arm and 694 in the simvastatin arm. There was no statistical difference between the groups, with a hazard ratio of 1.04 (95% CI: 0.94-1.15; p=0.46). Dr. Blazing did identify important limitations to the analysis, mainly related to the way diabetes was diagnosed (based on prescription of antidiabetic medication or two elevated fasting glucose levels rather than A1c or an OGTT). However, he shared the results of five sensitivity analyses using varying definitions and/or exclusion criteria, all of which were non-significant. As he noted during Q&A, the amount of incremental LDL-lowering produced by adding ezetimibe to simvastatin was about equivalent to the difference between a moderate and high dose statin. Studies comparing moderate and high dose statins have found an increase in diabetes risk of about 2/1,000 patient-years, and Dr. Blazing said that this study was well-powered enough to uncover that level of increased risk. Therefore, the results suggest that the increase in diabetes risk is more likely to be a statin-specific phenomenon rather than one related to LDL lowering more broadly; the ongoing PCSK9 inhibitor outcomes trials should also help shed much more light on this question.

Questions and Answers

Q: You mentioned that you would expect every 1 in 1,000 patients to be at risk. Was this study large enough to make a conclusion?

A: It could be underpowered but it’s bigger than any study done previously by far. It’s not as big as a meta-analysis but it’s about as big as you can get for an individual study.

Q: The risk is related to LDL according to Dr. Kastelein’s theory. The difference in LDL with ezetimibe is small. With a further reduction in LDL, would you expect an increased risk of diabetes?

A: Given the small change here, if it’s all LDL-related then no. If it has to do with the intensity of the statin, then possibly. Adding ezetimibe essentially turns the treatment into a high-dose statin, so you would expect to see 3/1,000 instead of 1. The trial was large enough to see that. Because of the number of cases and the power, we would have caught a high-intensity type effect.

Q: You had 1,400 new cases of diabetes. Is that figure high or low?

A: It was relatively high, in part due to the fact that the 1,400 cases for the primary analysis depended on the definition. The denominator was also lower because we excluded people missing data, about 800 from each arm. But in the sensitivity analyses we added them back and it didn’t affect outcomes.

Exhibit Hall

PCSK9 inhibitors were on full display in the exhibit hall, consistent with the interest they attracted throughout the conference. Amgen had the largest footprint with booths in both halves of the divided hall advertising the recently approved Repatha (evolocumab). In one of the more creative displays in the hall, the “PCSK9 Journey” display invited visitors to walk through a blown-up blood vessel that featured a full audiovisual tour. While Sanofi/Regneron’s Praluent (alirocumab) has not yet received EU approval, the companies still made their presence known with a large booth offering CV risk screenings for attendees and a wealth of information about the science of PCSK9 inhibition. While diabetes products occupied only a small portion of the exhibit hall, we were excited to see Merck’s booth featuring an entire wall devoted to the recent TECOS results for Januvia (sitagliptin) as well as head-to-head data vs. glimepiride. AZ, which had one of the most prominent booths in the hall, also dedicated a corner to its diabetes products, including touchscreens with key information.

 

-- by Emily Regier and Kelly Close