ESC 2017 (European Society of Cardiology)

August 26-30, 2017; Barcelona, Spain; Days #3-4 Highlights – Draft

ESC 2017 continued full steam ahead this week, following a packed weekend of learning (see our highlights coverage of days #1-2). On days #3-4, we saw new post-hoc analyses on Novo Nordisk’s LEADER (GLP-1 agonist Victoza), Lilly/BI’s EMPA-REG OUTCOME (SGLT-2 inhibitor Jardiance), AZ’s CVD-REAL study of all SGLT-2 inhibitor products vs. other glucose-lowering drugs, and Novo Nordisk’s SUSTAIN 6 (GLP-1 candidate semaglutide). Below these top four data-heavy highlights, we enclose insights from Drs. Neil Poulter, John Deanfield, Naveed Sattar, and David Fitchett on how cardiologists should be incorporating diabetes drugs into their clinical practice in this era of CVOTs. The experts weighed-in on how to personalize care for the type 2 diabetes patient with high CV risk – agents from two classes have now shown CV efficacy, but a GLP-1 agonist is perhaps advantageous for its weight loss effects, while an SGLT-2 inhibitor is perhaps best in the context of incipient heart failure. Let’s stop and celebrate for a moment that HCPs and patients even have a choice, with not one, not two, but three (!!!) glucose-lowering medicines on the market with cardioprotective benefits – Jardiance (empagliflozin), Victoza (liraglutide), and Invokana (canagliflozin) – of course, we realize that formularies etc play a bit role and ultimately determine choice. We were pleased to see standing-room-only crowds in these educational sessions. Cardiologists are showing keen interest in diabetes therapies with positive CV effects (as they should be), and we’re so glad this education is happening at a major conference like ESC, this year drawing >31,000 attendees from around the globe.

Read on for our top 10 highlights from days #3-4 of ESC 2017, also including learnings on PCSK9 inhibitors, DPP-4 inhibitors and differential impacts on heart failure, and how A1c independently affects CV risk. The talks at this meeting have been incredibly rich with commentary, and we’ll be expanding many them in our full report … stay tuned. If you have specific questions, of course let us know directly.

Top 10 Highlights

1. Dr. Neil Poulter presented new data from the LEADER trial, emphasizing that history of stroke/MI at baseline had no mediating effect in the significant association between GLP-1 agonist liraglutide (Novo Nordisk’s Victoza) and CV risk reduction. Simultaneous to the start of this session, Novo Nordisk issued a press release announcing these exciting post-hoc findings. This was a big deal.

2. Dr. Javed Butler presented a late-breaking post-hoc analysis of EMPA-REG OUTCOME, reporting consistent benefits to SGLT-2 inhibitor Jardiance (empagliflozin) across all categories of baseline heart failure risk. Again – wow.

3. Dr. Anna Norhammar presented late-breaking data from AZ’s CVD-REAL study, showing consistent benefits to SGLT-2 inhibitors vs. other glucose-lowering drugs on heart failure and all-cause death regardless of age, gender, CKD status, and concomitant medications. She concluded that this analysis adds to the robustness of initial CVD-REAL results, while also implying a potential benefit to this therapy class in a broader type 2 diabetes population. Kudos again to AZ for taking on this important work. “Real-world” has never been so important.

4. Dr. Esteban Jodar shared new insights on SUSTAIN 6, the pre-market CVOT for Novo Nordisk’s once-weekly GLP-1 agonist semaglutide. His slides featured extended Cox regression models to show that semaglutide’s CV benefits are sustained (no pun intended) throughout two years. This is a big drug coming – we love that the GLP-1 class keeps getting bigger and better.

5. Drs. Neil Poulter and John Deanfield led an interactive discussion of diabetes CVOTs and what cardiologists need to know. They acknowledged that cardiologists tend to be more comfortable prescribing oral agents rather than injectable ones, especially when it comes to medicines originating in other disease areas, but argued that it’s time to put a kibosh on this hesitation in the aftermath LEADER, which showed CV benefit to Novo Nordisk’s injectable GLP-1 agonist Victoza (liraglutide).

6. Drs. David Fitchett and Naveed Sattar shared practical advice for the cardiologist looking to prescribe an SGLT-2 inhibitor, answering key questions about renal safety, amputation risk, and the choice between GLP-1 and SGLT-2. This was an absolutely packed, standing-room-only session, an observation Dr. Sattar made to point out that SGLT-2 inhibitors are already of keen interest within cardiology settings (as they should be).

7. Dr. Darren McGuire criticized treatment guidelines that have assigned heart failure risk to all DPP-4 inhibitor products as a class effect. In a similar vein, he questioned the FDA’s recent decision to add heart failure warnings to the Januvia (Merck’s sitagliptin) and Tradjenta (Lilly/BI’s linagliptin) labels, calling it a “curious action.”

8. In an Amgen-sponsored lunch symposium, lipid experts Drs. Brian Ference, Stephen Nicholls, Kausik Ray, and Chris Packard attributed positive FOURIER results to the high margin of LDL reduction achieved by PCSK9 inhibitor Repatha (evolocumab). As such, they predicted similar findings to come from Sanofi/Regeneron’s ODYSSEY OUTCOMES study of Praluent (alirocumab), expected to read out in early 2018.

9. Dr. Lawrence Leiter made a two-pronged pitch for why cardiologists should care about glycemia: (i) better glycemic control has resulted in less CV disease, and (ii) as cardioprotective therapies help people live longer with diabetes, more microvascular complications of hyperglycemia can build up over the longer term.

10. Duke University’s Dr. Neha Pagidipati used data from Merck’s TECOS trial of DPP-4 inhibitor Januvia (sitagliptin) to probe for a correlation between BMI and CV outcomes in a diabetes patient population. She found that no class of overweight/obesity significantly affected CV risk, but that BMI <20 kg/m2 (indicating underweight) conferred heightened risk for adverse CV events.

Top 10 Highlights

1. LEADER Results Still Resoundingly Positive after Adjusting for Baseline History of Stroke/MI

Dr. Neil Poulter presented new data from the LEADER trial, emphasizing that history of stroke/MI at baseline had no mediating effect in the significant association between GLP-1 agonist liraglutide (Novo Nordisk’s Victoza) and CV risk reduction. Simultaneous to the start of this session, Novo Nordisk issued a press release announcing these exciting post-hoc findings. Across the entire CVOT (n=9,340), liraglutide reduced risk for the primary outcome of three-point MACE (non-fatal MI, non-fatal stroke, and CV death) by 13% compared to placebo (p=0.01 for superiority). Among 3,692 patients with a prior history of stroke/MI, the hazard ratio for three-point MACE was 0.84 in favor of the GLP-1 agonist (95% CI: 0.72-0.97), which meets threshold for statistical significance since the upper bound of the 95% confidence interval falls below 1.00. Among the 5,648 patients without prior stroke/MI, this hazard ratio for primary endpoint was 0.89 (95% CI: 0.76-1.05), with a trend in the right direction (numerically fewer MACE events on liraglutide vs. placebo), though this was not statistically significant. Importantly, the p-value for treatment interaction here was a non-significant 0.56, meaning history of stroke/MI did not significantly influence the overall results. Several presentations at ADA 2017 ruled out alternative explanations for Victoza’s CV benefit, reporting non-significant p-values for interaction on severe hypoglycemia, concomitant medication use, and recurrent CV events – none of these variables could explain the robust cardioprotection seen in LEADER, and now we can add history of stroke/MI to this list as well. We see a few key implications of the results Dr. Poulter described: First, as post-hoc analyses eliminate alternative explanations, evidence builds in support of CV benefit inherent to the liraglutide molecule, which could be important in convincing real-world HCPs to prescribe Victoza. Second, these findings add to our knowledge of what works and what doesn’t in CV risk mitigation, which bolsters our ability to approach diabetes management with cardioprotection in mind (this paradigm shift is already underway, but additional insights on CV risk reduction are always welcome). Third, diabetes CVOTs published to-date already enroll a relatively small slice of what is overall a highly heterogeneous type 2 diabetes population. The FDA recently approved a new CV indication for Victoza, but limited its applicability to people with type 2 diabetes and established CV disease (not to undercut how incredibly excited we were about this regulatory decision). It is reassuring to learn that liraglutide’s CV effects extended consistently across the risk spectrum of participants enrolled in LEADER at least – whether or not cardioprotection extends to lower-risk individuals will have to be determined through future research.

  • Dr. Poulter also broke down LEADER results for individual components of three-point and expanded MACE according to baseline history of stroke/MI. These are summarized in the table below.

Endpoint

Hazard Ratio for Participants with Previous Stroke/MI (95% CI)

Hazard Ratio for Participants without Previous of Stroke/MI (95% CI)

P-value for Interaction

Three-point MACE (non-fatal MI, non-fatal stroke, CV death)

0.84 (0.72-0.97)

0.89 (0.76-1.05)

0.56

Expanded MACE (non-fatal MI, non-fatal stroke, CV death, hospitalization for unstable angina, hospitalization for heart failure)

0.87 (0.77-0.98)

0.88 (0.78-1.00)

0.90

CV Death

0.80 (0.60-1.01)

0.76 (0.58-0.99)

0.79

MI

0.82 (0.66-1.01)

0.88 (0.70-1.11)

0.64

Stroke

0.92 (0.69-1.21)

0.80 (0.60-1.08)

0.52

Coronary Revascularization

0.89 (0.74-1.07)

0.93 (0.76-1.13)

0.75

Hospitalization for Unstable Angina Pectoris

1.15 (0.83-1.60)

0.76 (0.52-1.13)

0.12

Hospitalization for Heart Failure

0.80 (0.62-1.02)

0.95 (0.73-1.24)

0.33

Questions and Answers

Q: What is the place of compounds like empagliflozin and liraglutide that have shown CV benefit? Should they be moved to first-line therapy within treatment guidelines?

Dr. Poulter: In all fairness, this has been a huge sea change. There was uncertainty before with GLP-1 agonists, since lixisenatide came out with a neutral result (Sanofi’s ELIXA trial). I also admit that these trials were conducted in the context of patients with established CV disease, or at least very high CV risk, and medicine was given on top of standard of care. Based on these trials alone, I don’t believe we have sufficient evidence to support the use of these agents as first-line. On the other hand, the evidence for metformin as first-line treatment is also not great. Some guidelines are moving these drugs toward second-line.

Q: Have you stratified the results for patients with and without prior heart failure?

Dr. Poulter: We did. There was no statistical interaction, but the benefits of liraglutide were attenuated, though still in the right direction. Within LEADER, liraglutide reduced heart failure by 13% – again, this was not a statistically significant result, but it was in the right direction. This was reassuring following heart failure-related concerns for some of the DPP-4 inhibitors.

Q: Was the number of patients with heart failure at baseline relatively small?

Dr. Poulter: It was ~18%.

Q: Do you foresee GLP-1 agents being explored for heart failure in future trials?

Dr. Poulter: If you compare the benefits to heart failure we saw in EMPA-REG OUTCOME vs. LEADER, it really does look like the benefits of empagliflozin were driven by heart failure and CV death. There was blood pressure diuresis, there was rapid divergence of curves – all of this looks more beneficial from the heart failure point of view. The message from LEADER was that liraglutide looked safe from the heart failure point of view.

2. Jardiance (Empagliflozin) Lowers Risk for Heart Failure/CV Death in All Subgroups Stratified by Five-Year Heart Failure Risk

Dr. Javed Butler presented a late-breaking post-hoc analysis of EMPA-REG OUTCOME, reporting consistent benefits to SGLT-2 inhibitor Jardiance (empagliflozin) across all categories of baseline heart failure risk. The 6,314 participants in Lilly/BI’s large CVOT without a prior history of heart failure were stratified into (i) low-to-average, (ii) high, or (iii) very high risk groups for five-year likelihood of heart failure according to the Health ABC HF Risk Score (a well-validated assessment according to Dr. Butler). People in the lowest-risk category (~67%) experienced a 29% relative risk reduction with empagliflozin vs. placebo for the composite endpoint of heart failure hospitalization or CV death (HR=0.71, 95% CI: 0.52-0.96). Among those at low-to-average risk, the incidence of heart failure was 1.68/100 patient-years in the placebo group vs. 1.20/100 patient-years in the treatment group. As expected, this absolute risk increased for those categorized as high-risk for heart failure (~24%), with events occurring at an incident rate of 4.03/100 patient-years in the placebo arm vs. 2.07/100 patient-years in the empagliflozin arm. Still, empagliflozin conferred a significant CV benefit, in the form of 48% relative risk reduction for the heart failure/CV death composite (HR=0.52, 95% CI: 0.36-0.75). Moving up the scale to the very high-risk category (~5%), Dr. Butler reported heightened incidence of heart failure at 7.0/100 patient-years among placebo-treated participants vs. 3.8/100 patient-years among empagliflozin-treated participants. In this third category, Jardiance was associated with a 45% relative risk reduction for the heart failure/CV death composite (HR=0.55, 95% CI: 0.30-1.00). Due to fewer events in each subgroup compared to the overall EMPA-REG OUTCOME trial, it’s more challenging for any of these subgroup hazard ratios to meet criteria for statistical significance (hence the wide confidence intervals). That said, it’s quite notable that there was no significant interaction between baseline heart failure risk and Jardiance’s major CV benefits (p=0.428), because this underscores the drug’s potential cardioprotective value to a broader range of type 2 diabetes patients – in the real world, not all have established CV disease or a history of CV events, but these lower-risk people still stand to benefit from an agent effective in primary CV prevention. We’re impressed by the continued insights that have been gleaned from subsequent analyses of EMPA-REG OUTCOME, and this was no exception. Dr. Butler announced that these post-hoc findings are to be published soon in the European Heart Journal.

3. CVD-REAL Finds Consistent Benefits to SGLT-2 Inhibitor Treatment Regardless of Age, Gender, CKD, Concomitant Medications

Dr. Anna Norhammar presented late-breaking data from AZ’s CVD-REAL study, showing consistent benefits to SGLT-2 inhibitors vs. other glucose-lowering drugs on heart failure and all-cause death regardless of age, gender, CKD status, and concomitant medications. She concluded that this analysis adds to the robustness of initial CVD-REAL results, while also implying a potential benefit to this therapy class in a broader type 2 diabetes population. SGLT-2 inhibitors (whether AZ’s dapagliflozin – the most-prescribed SGLT-2 agent in Europe, J&J’s canagliflozin – most-prescribed in the US, or Lilly/BI’s empagliflozin) showed superiority over other glucose-lowering agents in reducing risk for heart failure regardless of age, with a hazard ratio point estimate of 0.58 (95% CI: 0.49-0.70) for those ≥65 years-old vs. 0.64 (95% CI: 0.44-0.93) for those <65 years-old (p=0.657 for interaction). This benefit was sustained for women (HR=0.54, 95% CI: 0.42-0.70) and men (HR=0.64, 95% CI: 0.54-0.76, p=0.275 for interaction), as well as for those with chronic kidney disease (CKD) at baseline (HR=0.46, 95% CI: 0.23-0.90) and those without (HR=0.63, 95% CI: 0.55-0.72, p=0.368 for interaction). Similarly, risk reduction for all-cause mortality was greater with SGLT-2 inhibitors vs. other diabetes drugs in both age groups (p=0.598), gender groups (p=0.971), and CKD status groups (p=0.097). SGLT-2 therapies were associated with superior risk reduction for a composite endpoint of heart failure or death in all subgroups: the p-value for interaction was a non-significant 0.76 for age, 0.452 for gender, and 0.76 for baseline CKD. As a reminder, relative risk reduction with SGLT-2 vs. other therapies in the overall CVD-REAL population was 39% for heart failure hospitalization (p<0.001), 51% for all-cause mortality (p<0.001), and 46% for the composite (p<0.001). Further detail on the consistency of effects seen with SGLT-2 inhibitors vs. other glucose-lowering drugs regardless of concomitant medications can be found below, but the key takeaway from this analysis is already apparent – the SGLT-2 class as a whole has demonstrated important benefits to heart failure and all-cause mortality in this real-world dataset, which includes people from a wide age range, from both genders, and with a variety of disease experiences in terms of comorbidities and other medications used. This latter point is particularly interesting, in our view, since randomized controlled trials of SGLT-2 inhibitors thus far have primarily evaluated CV effects in type 2 diabetes patients at high CV risk, which is only one slice of a heterogeneous type 2 population in real world. Dr. Norhammar called attention to the high number of primary prevention participants enrolled in DECLARE (AZ’s CVOT for dapagliflozin) compared to EMPA-REG OUTCOME (for Lilly/BI’s empagliflozin) or CANVAS (for J&J’s canagliflozin), suggesting that this upcoming RCT expected to complete in 2H18 will elucidate the impact of SGLT-2 inhibitors on lower CV risk patients. We look forward to this read out, since evidence from an RCT is much more likely to affect product labels and treatment guidelines (as opposed to an observational study like CVD-REAL, though we appreciate this tremendous effort/investment from AZ and the hint toward a cardioprotective class effect). We’d love to see SGLT-2 inhibitors indicated for CV protection in a broader type 2 diabetes population – the Jardiance (empagliflozin) label points to risk reduction for CV death among people with diabetes and established CV disease.

  • When adjusted for concomitant use of ACE/ARB agents, loop diuretics, beta blockers, aldosterone antagonists, and statins (from the CV area) and of insulin and sulfonylureas (from the diabetes area), CVD-REAL still showed a beneficial effect of SGLT-2 inhibitors on heart failure, all-cause death, and a composite of heart failure/all-cause death. On the composite endpoint, the p-value for interaction was a non-significant 0.606 for ACE/ARB agents, 0.076 for loop diuretics, 0.325 for beta blockers, 0.999 for aldosterone antagonists, 0.979 for statins, 0.369 for insulin, and o.791 for sulfonylureas. More specifically, the hazard ratio point estimate was 0.52 (95% CI: 0.44-0.63) among people on insulin vs. 0.58 (95% CI: 0.48-0.72) among people not taking insulin. For CVD-REAL participants taking sulfonylureas, the hazard ratio point estimate on this composite outcome was 0.55 (95% CI: 0.44-0.70) vs. 0.57 (95% CI: 0.50-0.66) for those not taking sulfonylureas. These findings underscore the consistency of SGLT-2 inhibitors in lowering risk for heart failure and death, despite background medications in any given patient’s treatment regimen. More often than not, real-world patients are on multiple therapies or are likely to switch therapies – after all, people don’t live in the controlled environment of an RCT – so we were pleased to see this focused analysis of CVD-REAL that took concomitant medications into consideration.
  • The positive results on SGLT-2 inhibitors in people with baseline CKD are intriguing in the context of recently-initiated outcomes trials from AZ and Lilly/BI of dapagliflozin and empagliflozin, respectively, in this patient population. The Dapa-CKD study for Farxiga was launched this past February, and is expected to complete in November 2020. Lilly/BI have yet to share timing for the Jardiance-in-CKD trial, but this clinical project was announced in June during ADA 2017. Data suggestive of renal protection has been a pleasant surprise finding from SGLT-2 inhibitor CVOTs so far, since all three commercial products in this class are actually contraindicated for people with renal disease. There’s certainly much more to unravel in this therapeutic area, and we’ll be keeping an eye out for updates on Dapa-CKD and on Lilly/BI’s planned trial.

4. SUSTAIN 6 Post-Hoc Analysis Underscores Durability of Effect

Dr. Esteban Jodar shared new insights on SUSTAIN 6, the pre-market CVOT for Novo Nordisk’s once-weekly GLP-1 agonist semaglutide. His slides featured extended Cox regression models to show that semaglutide’s CV benefits are sustained (no pun intended) throughout two years. Described another way, two-year graphs of three-point MACE (non-fatal MI, non-fatal stroke, and CV death), non-fatal MI, and non-fatal stroke all showed a consistent negative slope, designating relative risk reduction for the adverse CV outcomes with semaglutide vs. placebo at all points in time. As a reminder, the hazard ratio for these endpoints after two years was 0.74 for the primary outcome of three-point MACE (corresponding to a 26% relative risk reduction, p=0.02 for superiority), 0.74 for non-fatal MI (p=0.12), and 0.61 for non-fatal stroke (corresponding to a 39% relative risk reduction, p=0.04). Dr. Jodar explained how this constant benefit over time aligns with the hypothesis that semaglutide slows the progression of atherosclerosis, and thereby exhibits its cardioprotective properties – one possible explanation for the agent’s CV mechanism. The extended Cox regression line was near flat for CV death, and indeed, the hazard ratio for this component endpoint in SUSTAIN 6 was 0.98 (95% CI: 0.65-1.48). Dr. Jodar underscored that this was a short trial, designed to demonstrate non-inferiority for semaglutide vs. placebo. He pointed out that with longer-term follow-up, a relative risk reduction for CV death may have appeared. Moreover, that semaglutide demonstrated an advantage over placebo for atherosclerotic MACE events at every point in time suggests a durability of effect, according to Dr. Jodar. This supports Novo Nordisk’s plans for a longer, larger post-market CVOT. Semaglutide is currently under review with the FDA and EMA, and we eagerly await decisions expected in 4Q17.

Questions and Answers

Q: Can we say there is a class effect for the group of drugs known as GLP-1 analogs?

Dr. Jodar: For now, I don’t think we can say there is a class effect. We have four major CVOTs presented – two neutral, two positive. This is all very important and interesting. Probably, we are going to have a lot more data in the future. But for me, it’s pretty clear that we can’t argue for a class effect any longer.

Q: Can you say something about the A1c difference in the two SUSTAIN 6 treatment arms?

Dr. Jodar: All investigators were told to use local guidelines to manage their patients. With semaglutide, however, we are dealing with a very powerful drug, one that’s able to reduce blood glucose remarkably – the levels we’ve seen are incredible for something that’s not insulin. So yes, there were some differences in A1c between the groups, perhaps a big greater than other CVOTs, but that’s the reason.

5. SGLT-2 vs. GLP-1: What’s a Cardiologist to Do?

Drs. Neil Poulter and John Deanfield, both hailing from London, led an interactive discussion of diabetes CVOTs and what cardiologists need to know. The conversation steered toward how to choose between an SGLT-2 inhibitor (Lilly/BI’s Jardiance or J&J’s Invokana) and a GLP-1 agonist (Novo Nordisk’s Victoza) when both have shown cardioprotection in people with type 2 diabetes and high CV risk. The speakers acknowledged that cardiologists tend to be more comfortable prescribing oral agents rather than injectable ones, especially when it comes to medicines originating in other disease areas (Dr. Poulter suggested that subcutaneous injection reminds the average HCP of hypoglycemia, even if GLP-1 agonists aren’t associated with this risk). That said, they rejected this hesitation, encouraging cardiologists in the audience to eagerly embrace Victoza in their prescribing habits now that liraglutide has demonstrated a 13% risk reduction for three-point MACE and a 22% risk reduction for CV death in the LEADER trial. Moreover, the FDA and EMA have each added CV indications to the Victoza label. It will still take time and concerted education to get Victoza incorporated into regular cardiology practice, according to Dr. Poulter, but he asserted that providers in this field now have a responsibility to engage with all glucose-lowering, cardioprotective therapies. He explained that Victoza could be particularly beneficial for type 2 diabetes patients with high CV risk and overweight/obesity, as liraglutide and other GLP-1 agents are associated with greater weight loss vs. empagliflozin, canagliflozin, and SGLT-2 agents on the whole. On the other hand, Dr. Poulter pointed out that SGLT-2 inhibitors may be the preferred cardioprotective choice for diabetes patients with incipient heart failure, since this class appears to confer beneficial CV effects through non-atherosclerotic mechanisms. Dr. Deanfield echoed these clinical opinions, and added that cardiologists are going to have to get on board with more injectable therapies now that PCSK9 inhibitors are also making waves in CV medicine. He speculated that HCPs will start co-administering an SGLT-2 inhibitor alongside a GLP-1 agonist “quicker than you think,” even though no RCT has yet investigated additive CV efficacy. AZ’s DURATION-8 study of SGLT-2 inhibitor Farxiga (dapagliflozin)/GLP-1 agonist Bydureon (exenatide once-weekly) has shown superior A1c decline and weight loss with dual therapy vs. either monotherapy, and Dr. Deanfield pointed to a substantial real-world dataset to support safety of this combo regimen, since many diabetes doctors are prescribing both to the same patient. But no CVOT has evaluated SGLT-2/GLP-1 combination therapy (it’s very unlikely that AZ would sponsor such a study, since Bydureon reported neutral topline CVOT data – we’ll see the full EXSCEL results soon, at EASD – but we wonder if Lilly might eventually consider this with Jardiance and Trulicity).

  • Indeed, cardiologists are suddenly tasked with choice in diabetes therapy, but Drs. Poulter and Deanfield emphasized that these are very exciting decisions to be making. A couple years ago we didn’t have any diabetes drugs with CV benefit, and we now have three on the market. Dr. Poulter spoke to the “art of medicine,” and to the opportunity for personalized treatment regimens. We love the idea that diabetes patients can express preference for an oral or injectable, for a medicine more effective on blood pressure or body weight, and either way can pick up a prescription at the pharmacy that’s going to help prevent CV complications – we see this as an enormous win for the diabetes field, for the cardiology field, and for the patient community.

6. Drs. Fitchett, Sattar Answer Cardiologist Questions on SGLT-2 Class

Drs. David Fitchett and Naveed Sattar shared practical advice for the cardiologist looking to prescribe an SGLT-2 inhibitor, answering key questions about renal safety, amputation risk, and the choice between GLP-1 and SGLT-2. This was an absolutely packed, standing-room-only session, an observation Dr. Sattar made to point out that SGLT-2 inhibitors are already of keen interest within cardiology settings (as they should be). Dr. Fitchett explained that the contraindication/recommended dose reduction for all SGLT-2 inhibitor products as eGFR falls <60 ml/min/1.73m2 is not for safety concerns, but rather for a presumed attenuation of efficacy. This class of glucose-lowering agents works through the kidneys, and thus renal impairment seems incompatible with SGLT-2 therapy. That said, both EMPA-REG OUTCOME for Lilly/BI’s Jardiance (empagliflozin) and CANVAS for J&J’s Invokana (canagliflozin) showed impressive risk reduction for a composite renal endpoint, suggestive of renal-protective properties. In turn, Lilly/BI have announced plans to conduct an outcomes trial of Jardiance in people with chronic kidney disease (CKD), and AZ has initiated a similar study of its SGLT-2 inhibitor Farxiga (dapagliflozin) called Dapa-CKD. Dr. Fitchett hinted that diabetes patients with eGFR <60 ml/min/1.73m2 may actually stand to benefit the most from treatment with an SGLT-2 inhibitor, in terms of CV and renal outcomes alike – the truth will lie in outcomes data, and we look very forward to that.

  • In response to an audience question regarding canagliflozin and a signal for lower limb amputations, Dr. Fitchett mentioned that the background amputation rate – even for a diabetes patient population – is relatively low, though he acknowledged that this is a very visceral, undesirable complication (“amputations aren’t a major issue generally, unless you have one”). Furthermore, Dr. Fitchett emphasized that EMPA-REG OUTCOME showed no imbalance in amputations. He outlined two possibilities for this divergence in safety data between Invokana and Jardiance: (i) difference in molecule or (ii) difference in study population. He did not mention differences in trial design, and explained that despite amputations being collected retrospectively in EMPA-REG OUTCOME, he doesn’t believe any events were missed. This has been a complicated story to unravel, and our thoughts continue to evolve. The best-case scenario, in our view, will be a post-hoc analysis of CANVAS that elucidates driving factors for the amputation signal, variables that can be controlled in real-world clinical practice, or that at the very least can inform patient selection. We’ll have to wait-and-see. We do hope the CANVAS safety data prompts better patient education around foot care.
  • Other cardiologists in the audience wondered what factors they should consider in selecting a cardioprotective diabetes drug, especially in choosing between an SGLT-2 inhibitor and a GLP-1 agonist. Dr. Sattar chimed in about the value of oral dosing (for the SGLT-2 class) as compared to injectable dosing (for the GLP-1 class, at least for now), the former offering greater convenience and often being preferable to patients for that reason. He mentioned that the ongoing EMPEROR HF studies of empagliflozin in heart failure will collect quality of life outcomes (music to our ears!). Dr. Sattar emphasized the remarkable risk reduction for heart failure hospitalization found in EMPA-REG OUTCOME as well as CANVAS. While GLP-1 agonist Victoza (liraglutide) demonstrated non-inferiority on this endpoint in LEADER, Dr. Sattar implied that there’s very little chance GLP-1 agents will be investigated in dedicated heart failure studies (“it’s not going to happen”). Notably, this topic also came up during a similarly-structured, interactive session with Drs. Neil Poulter and John Deanfield (see above), who highlighted Victoza’s superior weight loss efficacy vs. SGLT-2 inhibitors. This clinical feature, they argued, makes the GLP-1 agonist a better choice for type 2 diabetes patients with high CV risk and overweight/obesity. From our point of view, it’s incredible that HCPs are faced with this choice in the first place, as this means we now have two diabetes therapy classes with demonstrated CV efficacy. We’re pleased that this will allow for more individualized treatment plans. And, we so appreciated that both sets of experts issued a resounding call-to-action for cardiologists to embrace the changing paradigm of treatment and to start prescribing these diabetes drugs, whether SGLT-2 inhibitors or GLP-1 agonists.

7. Dr. McGuire on the “Fake News” Behind DPP-4 Inhibitors and an Assumed Heart Failure Class Effect

Dr. Darren McGuire criticized treatment guidelines that have assigned heart failure risk to all DPP-4 inhibitor products as a class effect. In a similar vein, he questioned the FDA’s recent decision to add heart failure warnings to the Januvia (Merck’s sitagliptin) and Tradjenta (Lilly/BI’s linagliptin) labels, calling it a “curious action.” The TECOS trial for Januvia found a resoundingly neutral hazard ratio of 1.00 (95% CI: 0.80-1.20) for heart failure hospitalizations, with no imbalance between the sitagliptin and placebo arms. Two CVOTs for Tradjenta are ongoing: The CARMELINA study comparing linagliptin vs. placebo is expected to complete in December 2017, while the CAROLINA study comparing linagliptin vs. glimepiride (a sulfonylurea) is expected to complete in March 2019. Takeda’s EXAMINE trial for Nesina (alogliptin) reported a numerical imbalance in hospitalizations for heart failure, but this finding did not reach statistical significance (HR=1.22 in favor of placebo, p=0.22). Thus, the worry over heart failure/DPP-4 agents stems mainly from the SAVOR-TIMI study of AZ’s Onglyza (saxagliptin), which found an unexpected 27% increase in risk for this safety endpoint (p=0.007). That said, Dr. McGuire also reviewed a sub-analysis of EXAMINE, dividing participants by whether or not they had a baseline history of heart failure: The hazard ratio for heart failure hospitalization in the trial was 1.00 for people with a prior event (p=0.99) vs. 1.76 for people without a prior event (p=0.026). The p-value for this interaction was just above 0.06, not quite statistically significant, though it did approach that 0.05 threshold. Dr. McGuire argued that regulatory decisions, like clinical practice decisions, should be evidence-based – to this end, he underscored the robust safety data on Merck’s Januvia, published in peer-reviewed literature. We echo his surprise at the FDA’s label updates for Januvia and Tradjenta. While we certainly don’t want to undercut the importance of drug safety, and while we appreciate all FDA does on this front (with limited resources at that), we’d also hate to see providers discouraged from using Januvia or Tradjenta in patients for whom they are safe and for whom they could be incredibly beneficial. DPP-4 inhibitors have shown CV safety on three-point MACE, have demonstrated glucose-lowering efficacy, and boast a long history of safety/tolerability in the field. Thought leaders, including Dr. Robert Ratner, have touted the advantages of this therapy class for the elderly, for patients with renal impairment, and for any diabetes patient earlier in the course of disease progression. Dr. McGuire concluded his talk with a list of key questions that need to be addressed – hopefully these answers can put the DPP-4/heart failure issue to bed, so that safe/effective therapies are maximally accessible.

  • Are these drug effects, or class effects?
  • Are they patient-specific effects? Can we identify patients at-risk to inform patient selection?
  • What’s the mechanism for saxagliptin/increase in heart failure?

8. Lipid Experts Sound Off on FOURIER, ODYSSEY OUTCOMES

In an Amgen-sponsored lunch symposium, lipid experts Drs. Brian Ference, Stephen Nicholls, Kausik Ray, and Chris Packard attributed positive FOURIER results to the high margin of LDL reduction achieved by PCSK9 inhibitor Repatha (evolocumab). As such, they predicted similar findings to come from Sanofi/Regeneron’s ODYSSEY OUTCOMES study of Praluent (alirocumab), expected to read out in early 2018. Participants in FOURIER were followed for one-two years (median follow-up 26 months), and LDL dropped by a mean 59% on Repatha (from an average 92 mg/dl to 30 mg/dl). The relative risk reduction for the primary composite endpoint (CV death, non-fatal MI, non-fatal stroke, hospitalization for unstable angina, or coronary revascularization) was 15% overall (HR=0.85, p<0.0001). Dr. Ference reported a hazard ratio of 0.83 (95% CI: 0.77-0.90) when the data is adjusted for duration of evolocumab treatment, which corresponds to 17% relative risk reduction. ODYSSEY OUTCOMES for alirocumab is a longer trial, following patients for two-three years instead of one-two, and Dr. Ference thus forecasted a slightly lower hazard ratio of 0.80 (95% CI: 0.77-0.83) for the same primary endpoint. He postulated that if statins trials were run for much longer, to the point where LDL dropped to a similar extent as in PCSK9 trials, we would see on par risk reduction for adverse CV outcomes. In fact, he set the stage for this symposium by establishing the link between LDL-lowering and proportional CV risk reduction – this holds true across a diverse array of clinical trials, meaning the mechanism of LDL-lowering doesn’t matter as much as the extent and efficiency of cholesterol reduction. This strong correlation indicates that any agent able to aggressively lower LDL, including Sanofi/Regeneron’s PCSK9 inhibitor Praluent, would be expected to also show a meaningful decline in CV events – hence, the suggestion that we’ll have a second positive PCSK9 CVOT before too long. This optimism for ODYSSEY OUTCOMES (at an Amgen-sponsored session, no less) was exciting to hear. Dr. Packard expressed a positive outlook for the field of lipid management. He described an unmet medical need – at the population level, people are failing to reach LDL goals with statins alone – that can now be filled by PCSK9 inhibitors. We have good reason to believe in lowering LDL to prevent CV disease, and we now have a way to achieve the necessary LDL-lowering for almost all patients (>95% of participants in clinical studies of a PCSK9 inhibitor have reached or surpassed target LDL) – Dr. Packard called this a “fortunate situation.” We’re then left with the unfortunate situation of very poor reimbursement for these medicines. According to a National Lipid Association (NLA) survey, prior authorizations for a PCSK9 inhibitor are denied 96% of the time (this is scary high). ODYSSEY OUTCOMES could be a light at the end of the tunnel, as having two positive PCSK9 inhibitor trials showing a decrease in CV events could sway payers in the right direction (preventing heart attacks and hospitalizations holds tremendous cost-saving potential).

  • Dr. Nicholl’s remarks were focused on how PCSK9 inhibitors affect plaques, promoting a regression of atherosclerosis. This is a clear mechanism for CV risk reduction associated with PCSK9 agents. Given the hypotheses circulating in the diabetes field that GLP-1 agonists (namely, Novo Nordisk’s liraglutide and semaglutide) may also be cardioprotective due to atherosclerotic effects, we’re curious about the combined CV benefits that might be achieved through PCSK9/GLP-1 co-administration. On the other hand, SGLT-2 inhibitors (namely, Lilly/BI’s empagliflozin and J&J’s canagliflozin) appear to offer CV benefits through a non-atherosclerotic mechanism, since the overall risk reduction with these agents seems to be driven more so by heart failure and CV death vs. stroke and MI. PCSK9/SGLT-2 co-administration might then cover more bases in improving CV outcomes for people with diabetes, and possibly comorbid dyslipidemia (but even without the comorbidity, thought leaders at this meeting have emphasized the power of lipid-lowering in best practice diabetes care). This is all speculation for now, but we’d love to learn more about these therapeutic possibilities. As attendees were finishing their sandwiches, Dr. Ray concluded this lunch symposium with a list of lingering questions for the PCSK9 inhibitor class, one of which was “how can these agents be applied in specific populations?” – like diabetes.

9. Dr. Leiter’s Call to Cardiologists: Attend to Glycemic Control

Dr. Lawrence Leiter made a two-pronged pitch for why cardiologists should care about glycemia: (i) better glycemic control has resulted in less CV disease, and (ii) as cardioprotective therapies help people live longer with diabetes, more microvascular complications of hyperglycemia can build up over the longer term. The link between A1c and microvascular complications is uncontroversial – Dr. Leiter briefly summarized findings from the DCCT/EDIC to underline this fact. He acknowledged more confusing evidence from individual studies of how glycemic control impacts macrovascular disease, but presented a meta-analysis of four landmark trials – UKPDS, ACCORD, ADVANCE, and VADT – that did show a modest but significant 9% risk reduction for major CV events with improved glycemic control. The same meta-analysis calculated a 15% risk reduction for MI specifically, which Dr. Leiter described as clinically-meaningful. He alluded to a more recent study, just published in Diabetes, Obesity, and Metabolism in June 2017, that analyzes diabetes CVOTs completed to-date, and finds a greater magnitude of CV risk reduction with greater  estimated differential glycemic exposure (% A1c x years) between treatment groups. This is an interesting conclusion, since diabetes CVOTs are designed with glycemic equipoise in mind, so that CV effects can be linked to the specific molecule rather than to general glucose-lowering. Dr. Leiter’s main takeaway from this discussion, however, was not to undermine the exciting positive data we have on liraglutide, semaglutide, empagliflozin, and canagliflozin, but instead to emphasize that cardiologists shouldn’t ignore glycemia itself, even with the advent of cardioprotective diabetes drugs. He shared one graph of diabetes complication rates between 1990 and 2010: While MI and stroke were on the decline during these two decades, the incidence of end-stage renal disease was flat for the diabetes patient population overall, and actually rose among older people with diabetes. “This is not surprising,” Dr. Leiter articulated, “because if people aren’t dying of CV disease, they’re living long enough to encounter other complications of diabetes.” We appreciated this approach to convince cardiologists to attend to glycemia in their patients, especially since a majority of sessions at this meeting (and at many recent conferences, for that matter) have focused primarily on getting cardiologists to prescribe diabetes drugs once they have yielded positive CVOT data.

10. The Obesity Paradox: Does It Improve CV Outcomes in People with Diabetes?

Duke University’s Dr. Neha Pagidipati used data from Merck’s TECOS trial of DPP-4 inhibitor Januvia (sitagliptin) to probe for a correlation between BMI and CV outcomes in a diabetes patient population. She found that no class of overweight/obesity significantly affected CV risk, but that BMI <20 kg/m2 (indicating underweight) conferred heightened risk for adverse CV events. The hazard ratio for four-point MACE (non-fatal MI, non-fatal stroke, CV death, or hospitalization for unstable angina) was 1.87 (95% CI: 1.23-2.84) for TECOS participants who were underweight vs. those who were of a healthy weight, after adjusting for age, sex, race, history of CV disease, and other variables. The 95% confidence interval for four-point MACE hazard ratio crossed the line of unity for all other BMI subgroups: Among those with overweight (BMI 25-29.9 kg/m2) vs. healthy weight, the adjusted hazard ratio for four-point MACE was 0.96 (95% CI: 0.82-1.12). Among those with class I obesity (BMI 30-34.9 kg/m2) this value was 0.89 (95% CI: 0.76-1.05), among those with class II obesity (BMI 35-39.9 kg/m2) it was 0.97 (95% CI: 0.80-1.19), and finally, among those with class III obesity (BMI ≥40 kg/m2) it was 0.99 (95% CI: 0.76-1.28). Dr. Pagidipati speculated that having type 2 diabetes and being underweight perhaps indicates some other underlying disease, which is one possible explanation for the significant association with adverse CV outcomes. An important theme that arose during Q&A is how this analysis contributes to the controversy surrounding the obesity paradox, or the notion that obesity is sometimes associated with lower CV risk – Dr. Pagidipati pointed out that this paradox is especially murky in the context of diabetes. She summarized, “here, we don’t find evidence for the paradox, but we also don’t find that it’s necessarily worse to have overweight/obesity.” These findings were not meant to make any statement on weight recommendations in clinical practice, Dr. Pagidipati underscored, but rather, they serve as a launching pad for more dedicated inquiries into the relationship between obesity, diabetes, and CV disease. We agree that there’s a disappointing lack of research in this area, and we were glad to see the topic raised.

 

-- by Payal Marathe and Kelly Close