ESC 2019 (European Society of Cardiology)

August 31 – September 4, 2019; Paris, France; Days #1-3 Highlights – Draft

Executive Highlights

  • Full results from the DAPA-HF trial of Farxiga in HFrEF patients (both with and without diabetes) have stolen the show so far, revealing highly statistically significant results on reducing worsening heart failure events and CV death (26% RRR; p=0.00001). Farxiga also met significance on secondary outcomes of CV death or HF hospitalization (25% RRR; p=0.00002), total HF hospitalizations and CV death (25% RRR; p=0.0002), and all-cause death (17% RRR; p=0.022). Also striking was the very clean side effect profile. Reflecting the landmark nature of these results and the palpable excitement from the audience of thousands of cardiologists, loud cheers and applause emerged as these impressive results were revealed (watch this on the ESC website, where the video and slides are free for 120 days). Farxiga now has robust data indicating it as a potential treatment for both prevention and treatment of heart failure—wow!

  • Equally notably, and with exquisite timing, ESC released its new guidelines in diabetes, prediabetes, and CVD over the weekend, with a corresponding presentation Monday afternoon. It’s a landmark update, in our view – the first ever guidelines to recommend using SGLT-2 inhibitors and GLP-1 agonists as first-line therapies in patients with established ASCVD or at very high CV risk. There’s some debate on this point, but we’re glad to see such a progressive take from ESC/EASD – see KOL reactions below as well. The guidelines also move past the idea of “primary vs. secondary prevention” patients in terms of risk stratification and roll out a new set of risk categories for better stratification.

    • From our view, we see the technical reasons behind the differences between as GLP-1 and SGLT-2 inhibitors and metformin – the same benefit and “level of evidence” isn’t there for metformin, nor is there a manufacturer who would invest in such trials. We believe that on balance, if HCPs understand well and have time to sort out metformin titration, metformin of course has benefits, though not the same level of cardioprotection as GLP-1 and SGLT-2 inhibitors. Presumably, it will be straightforward to work out risk stratification for the newer therapies, though deciding between them (or where both might be suitable) may take some time to work out.

  • Full results from PARAGON-HF of Novartis’ Entresto in HFpEF were also presented, revealinig a narrow miss of superiority on a primary endpoint of composite heart failure hospitalization and CV death (p=0.059), as we reported a couple of weeks ago (Novartis’ HF therapy Entresto narrowly misses superiority in PARAGON-HF CVOT – August 2, 2019). These were disappointing results for Novartis since otherwise Entresto would certainly have been positioned as the first therapy for heart failure. Full results were simultaneously published in NEJM – see here. Indeed, positive results in PARAGON-HF would have represented the first ever for a potential treatment of HFpEF. We note that SGLT-2 inhibitors are also under investigation specifically in this population of people with HFpEF (EMPEROR-Preserved for Lilly/BI’s Jardiance, DELIVER for AZ’s Farxiga), contributing to a recent push to help fill this vast unmet need – certainly based on DAPA-HF, potential for dapagliflozin and potentially others in the class in the future seems very strong.

Greetings from Paris, where ESC 2019 is off to a stellar start. See below for our top highlights from the first three days of the meeting, and be sure to see our preview for what’s to come in future installments!

Top Seven Highlights

1. DAPA-HF Full Results: Highly Significant 26% RRR on CV Death/HHF in HFrEF Patients; Consistent Effect in Diabetes and Non-Diabetes Subgroups; Trial Enrolled 45% w/ Type 2 and 37% with Prediabetes

To an enthusiastic standing-room only crowd (watch the video here and download slides – as we understand it from ESC, it’s free for 120 days), University of Glasgow’s Dr. John McMurray revealed highly positive results from the DAPA-HF trial, where dapagliflozin demonstrated a robust effect in HFrEF patients (both with and without diabetes) on reducing worsening heart failure events and cardiovascular death. See AZ’s press release on its global side regarding the full results here – note that the US site has a slightly different press release noting that the drug is not approved for type 1. On the primary composite outcome of CV death/HF hospitalization/urgent HF visit, dapagliflozin 10 mg treatment delivered a 26% relative risk reduction when compared to placebo (HR=0.74; 95% CI: 0.65-0.85; p=0.00001; NNT=21) – results were even more pronounced in men and in those with BMI over 30. Reflecting the landmark nature of these results and the palpable excitement from the audience of thousands of cardiologists, loud cheers and applause emerged as these impressive results were revealed. Dapagliflozin treatment also met significance on each of the separate components of the primary outcome (worsening HF event and CV death, see below). Very notably, treatment effect was consistent across both patient populations of people with diabetes and people without diabetes—Dr. McMurray characterized this as an “entirely consistent” benefit in people with and without diabetes. On this point, we do note that the large majority (67%) of participants without diabetes in the trial did have prediabetes – we’re excited about what DAPA-HF may mean in this regard for improving outcomes in people with prediabetes as well. Dapagliflozin is now the first SGLT-2 inhibitor to show utility as a treatment for heart failure and the first SGLT-2 inhibitor to demonstrate in an RCT beneficial effects in people without diabetes. Wow!

  • Dapagliflozin treatment met significance on prespecified secondary outcomes of (i) CV death or HF hospitalization (HR=0.75; 95% CI: 0.65-0.85; p=0.00002); (ii) total HF hospitalizations and CV death (RR=0.75; 95% CI:0.65-0.88; p=0.0002); and (iii) patient reported outcomes on the Kansas City Cardiomyopathy Questionnaire (2.8 point difference, p<0.001). Dapagliflozin treatment also nominally met significance on all-cause death (HR=0.83; 95% CI: 0.71-0.97; p=0.022). These are highly robust data, indicating clinically meaningful reductions in important outcomes for providers and patients.

  • Dr. McMurray highlighted that dapagliflozin “does everything we want a new heart failure drug to do: (i) reduce hospital admissions; (ii) increase survival; and (iii) improve symptoms – the three key objectives of treatment.” And to boot, these effects with dapagliflozin are on top of “excellent existing therapy,” such as Novartis’ blockbuster HF therapy Entresto (sacubitril/valsartan). Might the field eventually see combination therapies of dapagliflozin with various other heart failure therapies? Could we see polypills of multiple heart failure therapies include dapagliflozin – and how might these multi-therapy treatments fit into treatment algorithms that increasingly emphasize personalization of treatment to each patient? What might a polypill of Entresto and Farxiga look like?

  • University of Brescia’s Dr. Marco Metra provided the independent commentary for DAPA-HF, highlighting that these results mean that “we are facing a new era in heart failure treatment.” Dr. Metra emphasized the landmark nature of DAPA-HF, noting that dapagliflozin is “acting with a completely new mechanism of action” and providing highly clinically meaningful benefits on heart failure outcomes. Our sense is that many cardiologists feel the same way – a quick search on social media found several cardiologists enthusiastically sounding off on DAPA-HF results, noting its small NNT (21) over a short period of time as a potentially revolutionary and paradigm-shifting treatment in clinical practice. As a sidenote – many research-scientists in cardiology are very active on social media and ESC does a great deal to encourage this, which is impressive as it enables much greater and faster sharing of results among a much bigger field.

    • See the tweet here from Stephan Achenbach that shared the audience applause when the first Kaplan-Meier curve was shared on the primary compositive outcome

  • With these positive results, dapagliflozin becomes the first SGLT-2 inhibitor to demonstrate efficacy in people without diabetes. Farxiga is currently only indicated for the treatment of type 2 diabetes, but we would hopefully see a label expansion to reflect DAPA-HF results for Farxiga’s use in HFrEF patient—both with and without diabetes, pending regulatory approval, of course.

  • This is a huge step for AZ and the broader SGLT-2 class, and we’re excited about the potential for these therapies to impact the large intersection of heart failure patients with prediabetes that do not have type 2 diabetes. As a reminder, 82% of participants in DAPA-HF had either diabetes or prediabetes. We’re hopeful that expanded use of Farxiga in this patient population of those with HFrEF and prediabetes could have positive impacts on glycemia that lead to prevention of type 2 diabetes as well. For more, see our landscapes for SGLT-2 inhibitors in heart failure here and SGLT-2 inhibitors in CKD here.

  • What’s the next step on the global regulatory front? A request for an expanded label would hopefully be made soon to CDER, presumably, although we are not sure which division AstraZeneca would submit to since there is not currently a pathway for pre-diabetes as we understand it. Importantly, there were those in the trial who saw a benefit (see below) who had neither diabetes nor pre-diabetes. Perhaps the submission would be to the Division of Cardiovascular and Renal Products (DCaRP) rather than to the Division of Endocrinology and Metabolic Products (DMEP).

  • Looking ahead, Farxiga is also being studied in HFpEF in the DELIVER trial. DELIVER (n=4,700) aims to examine the same primary endpoints of CV death and worsening of HF as DAPA-HF, but in patients with HFpEF, and is expected to complete in June 2021. Very notably, HFpEF is the form of heart failure thought to be more tightly linked to adiposity, obesity, and diabetes; moreover, current clinical evidence suggests that 30%-40% of people with HFpEF also have type 2 diabetes. There are no currently approved treatment options for HFpEF – Novartis’ Entresto (currently only indicated for HFrEF) just missed superiority in its HFpEF trial, PARAGON-HF (see below). Lilly/BI’s Jardiance is also being studied in HFpEF in the EMPEROR-Preserved trial.

  • In a conversation with our team, AZ’s Dr. Naeem Khan (VP, Medical, US CVMD) and Kiersten Combs (VP, US CVMD) highlighted these findings as “groundbreaking” and a “breakthrough in the heart failure space.” Dr. Khan noted that “We have a possible label for prevention [of heart failure], and now, with these data, treatment. For diabetes and for any HF patient – now that is groundbreaking. The possibility it has for clinical practice and for patients is what makes me really excited. If you add dapagliflozin on top of conventional therapy, you get a significant 26% reduction (p=0.0001) — we have the potential to decrease events, decrease mortality, and improve quality of life. I’m excited!” Dr. Khan also added that from the cardiologist perspective, he was very excited for the possibilities that dapagliflozin could provide for both physicians and patients. “It’s a paradigm shift from what we’ve known in the past. In the study, we had patients who were very well-controlled with the current standard of care but still had great benefit. As you go through each hierarchical analysis, you see the benefits there: for HF, CV death, and total mortality. It is startling what we can now provide to patients. It could be a paradigm shift for heart failure.”
  • Trial design and study population: The DAPA-HF trial was an international, multi-center, parallel group, event-driven, randomized, double-blind, placebo-controlled study (n=4,744) in adults with HFrEF (NYHA functional class II-IV, LVEF ≤40%) for at least two months. Importantly, 45% of study participants had known or undiagnosed diabetes (n=2,137), 37% had prediabetes (n=1,750) and 18% had normoglycemia (n=857). Participants were expected to receive background standard of care treatment, such as ACE inhibitors, ARBs, beta-blockers, MRAs, or neprilysin inhibitors. Among other factors, patients with type 1 diabetes were excluded. After meeting participation criteria, participants were randomized to receive either Farxiga (10 mg/day) or a placebo. The primary outcome for the study was time to cardiovascular death or the worsening of heart failure (defined as hospitalization or an urgent heart failure visit). Secondary outcomes also included renal measures, including time to (i) ≥50% sustained decline in eGFR; (ii) reaching ESRD; or (ii) renal death.

Close Concerns Questions:

  • Will SGLT-2 inhibitors now receive more consideration to become standard-of-care in heart failure patients, regardless of diabetes status? As a reminder, current SOC for HFrEF includes ACE inhibitors, ARBs, beta blockers, and mineralocorticoid-receptor antagonists.

  • Could a dedicated heart failure indication for Farxiga pave the way for broader use of this therapy in people with prediabetes? Could such usage help to bend the curve toward greater prevention of diabetes?

  • How will AZ shift marketing and commercialization of Farxiga to reflect this new patient population that may benefit?

  • Will data on glucose-lowering effects in those with prediabetes be released?

2. ESC Releases 2019 Guidelines on Diabetes, Prediabetes, and CVD: Recommends First-Line SGLT-2/GLP-1 Treatment in Patients with High Risk for or Established CVD

For the first time since 2013, ESC has updated its diabetes, prediabetes, and CVD guidelines – and wow is it a big update! See the full publication here. There’s so much to unpack here, but most notably, these new guidelines generated lots of discussion regarding their recommendation of using SGLT-2 inhibitors and GLP-1 agonists as first-line therapies in patients with established ASCVD or at very high CV risk. This represents the first time a major guideline has recommend first line usage of a non-metformin therapy in type 2 diabetes—a huge paradigm shift and landmark moment in the field. To be sure, this decision generated some debate particularly among endocrinologists – see our KOL reactions below, some of which focused on this point – but we remain generally positive about this move since we doubt there is any worry that doctors who know how to use metformin well will do so as a concomitant therapy – and perhaps some who don’t use it so well will shift, which would be positive for those that do not have HCPs that know how to titrate metformin. Any shift to have patients on these more efficacious therapies is a win in our minds, and we’re hopeful about how these guidelines may help to combat issues of therapeutic inertia that so often hinder optimal use of these therapies. Presumably a number of patients who are on metformin do not have therapies added fast enough and that should be less of a problem for those that go onto GLP-1 or SGLT-2s first though we’d of course want to make sure the segmentation worked well so that those who could benefit from the new guidelines are the ones that do. See below for top highlights from the detailed presentation on these guidelines, including (i) Dr. Francesco Cosentino’s broad overview; (ii) Dr. Maddalena Lettino on the management of dyslipidemia; (iii) Dr. Nikolaus Marx on how CVOT data was incorporated; and (iv) Dr. Petar Steferovic on heart failure considerations. And first, see the much-discussed treatment algorithm right below.

ESC 2019 Guidelines Type 2 Diabetes Treatment Algorithm:

Dr. Francesco Cosentino kicked off of the packed-to-capacity session with a broad overview of what’s new in the 2019 ESC guidelines. Dr. Cosentino underscored that the CVOT era of diabetes research – specifically the last five years – has been “the most exciting time in diabetes research ever.” As a result, there’s been an “unprecedented increase in the amount of evidence for diabetes therapies,” which is clearly evident in these new guidelines. Time and time again, Dr. Cosentino circled back to just how influential these “stunning results” from CVOTs have been in shifting the paradigm in diabetes to improved treatment and prevention of CVD. Dr. Cosentino emphasized two major updates in these guidelines of note:

  • New CV risk categories in patients with diabetes: Dr. Cosentino noted that these new risk categories are heavily impacting ESC’s new recommendations, and he is anticipating that the new risk categories will provide proper diagnostic and therapeutic approaches to patient care. Notably, this is a move away from the binary distinction of “primary vs. secondary” prevention patients and move toward better considering CV risk on a wider spectrum. See table below for the new risk stratification categories.

    • We note several questions regarding risk stratification for people with type 1 diabetes:

      • Those who are diagnosed aged 10 and under are clearly at very highest risk according to the new guidelines.

      • Those who are under 35 that were diagnosed ten or fewer years before are at moderate risk according to the new guidelines.

      • This seems to imply that those aged 11-35 with more than 10 years duration of diabetes and no target organ damage are at high risk.

      • It is not clear whether those 35 and older with more than 10 years duration of diabetes are at “High Risk” or “Very High Risk” and what other risk factors this depends upon. Those 35 and older with target organ damage and multiple risk factors are also presumably at Very High Risk – if they are also type 1 diagnosed before age 10, we are not sure if the risk stratification is even higher.

      • It is not clear what risk bucket these groups of people with type 1 fall into:

        • Over 35 and onset after aged 10 with no target organ damage or any major risk factors (presumably High Risk?)

        • Over 35 and onset after aged 10 with no target organ damage and with one of two major risk factors (presumably High Risk?)

      • While for GLP-1 use, this is a moot point since they are not currently approved for type 1, Novo Nordisk has previously said it may initiate trials in type 1 if oral semaglutide is approved. While for SGLT-2 use, this is a moot point in the US, SGLT-2s for type 1 are currently approved in the EU and Japan. There have been no outcomes trials for type 1 although many more people with type 1 are reaching advanced ages and evidence on cardiovascular disease is extremely sparse – we advocate for changes on this front now that there are approved therapies in the field and particularly given the magnitude of positive data for SGLT-2s for heart failure and CVD and renal disease.

Very High Risk

Patients with diabetes and established CVD

Or other target organ damage

Or three or more major risk factors

Or early onset type 1 diabetes of long duration (>20 years with early onset defined as diagnosis of aged 10 and younger)


Patients with diabetes duration ≥10 years without target organ damage plus any other additional risk factors – presumably this is both type 1 and type 2


Young patients (type 1 <35 years old or type 2 <50 years old) with diabetes duration <10 years, without any other risk factors


There are no patients with diabetes in this segment according to ESC.

  • First-line use of cardioprotective therapies: Dr. Cosentino asserted that CVOT data now allows the community to “start treatment with drugs that have proved cardioprotective effect” and that we no longer have to “indulge with starting with metformin—we now have evidence that we should start right away with these drugs [SGLT-2/GLP-1].” Much more on this below in our KOL reactions highlight. Presumably it doesn’t have to be “either/or” – there doesn’t seem to be much wrong with starting metformin and an advanced therapy at the same time, as long as titration is well understood and there are not titration problems that cause adherence issues (this is also true for advanced therapy, such as nausea associated with GLP-1).

Dr. Maddalena Lettino presented on the management of dyslipidemia in these new guidelines, emphasizing that statins should be prescribed on an individual basis and PCSK9 inhibitors should be prescribed if other therapies fail – overall we perceived more positivity around PCSK9s than has been articulated earlier. She emphasized that the aforementioned revised CV risk categories are important here in evaluating and stratifying patients to promote personalized goals. For example, LDL-C targets range from <100 mg/dL to <55 mg/dL for patients with moderate to very high risk, respectively – previously this was <100 mg/dL to <70 mg/dL. Under the new guidelines that target at least a 50% reduction in LDL-C for very high-risk patients, Dr. Lettino stated that this patient group should achieve lower LDL-C targets than those at high or moderate risk.  

  • Key treatment recommendations heavily drew on data from the IMPROVE-IT, FOURIER, and ODYSSEY trials, and the totality in evidence made it clear that (i) statins, the state-of-the-are therapy, prevent CV events and are also related to adverse events in patients with diabetes and (ii) ezetimibe or PCSK9 inhibitors should be used either on top of a statin (or alone in statin intolerant patients) to reduce LDL-C. Broadly, as noted, we see these guidelines as being very positive toward the usage of PCSK9 inhibitors and more so than previously seen with other professional societies. This is likely since they have only been approved for a couple of years – but we would anticipate the guidelines having a major positive impact on the companies that promote them, Sanofi/Regeneron and Amgen, particularly given the price reductions in the US.

Dr. Nikolaus Marx gave a comprehensive lesson on the current CVOT landscape, as well as the changes to the ESC treatment algorithm that reflect these results. See slides here. Overall, he stressed the guidelines’ shift towards a more evidence-based and individualized approach to treating diabetes, based on the current portfolio of completed CVOTs. Most notably, the plethora of CVOT data has largely reshaped the ESC’s guidelines to start with considering CV risk rather than A1c when prescribing drugs. For drug naïve patients, guidelines recommend first determining whether a patient has ASCVD or high/very high CV risk. If the patient does, then a GLP-1 or SGLT-2 is recommended based on CVOT data. If the patient does not have ASCVD or is at moderate risk, metformin is instead recommended based on UKPDS. Similarly, if a patient is already on metformin, use an SGLT-2 or GLP-1 if the patient has ASCVD or high/very high CV risk. After this preliminary funneling, A1c comes into play.

The entire SGLT-2 class (empagliflozin, canagliflozin, and dapagliflozin) is recommended for reduction in CV events for patients with type 2 diabetes and CVD, while empagliflozin is also recommended to reduce the risk of death based on its all-cause mortality benefit. Liraglutide, semaglutide, and dulaglutide are also recommended for reduction in CV events and liraglutide for reduction in risk of death based on its all-cause mortality data. For DPP-4 inhibitors, saxagliptin is not recommended for patients with type 2 diabetes at a high risk of heart failure due to the risk signal seen in SAVOR-TIMI

  • As a review, CV support for metformin comes from the UKPDS study, which showed a 39% relative reduction in myocardial infarction in a subgroup of overweight patients without CV disease vs. conventional therapy. In terms of DPP-4 inhibitors, all compounds within the class showed CV safety, but none have shown superiority. Notably, saxagliptin also showed an increased risk of hospitalization for heart failure in the SAVOR-TIMI trial. Looking at SGLT-2 inhibitors, all agents reduced hospitalization for heart failure and CV death (though dapagliflozin uniquely did not meet significance on 3P-MACE). Empagliflozin also reduced all-cause mortality. The GLP-1 class has produced more mixed results. Lixisenatide and exenatide have confirmed CV safety but have not shown reduction in CV events (though exenatide just missed it, and general consensus appears to suggest this is because of the much broader group in that trial). On the other hand, LEADER, SUSTAIN-6, HARMONY, REWIND, and PIONEER 6 all showed reductions in 3P-MACE, with liraglutide also reduced all-cause mortality in LEADER.

Finally, Dr. Petar Steferovic presented the 2019 guidelines’ strong take on glucose lowering drugs and heart failure, starting off his talk with a reminder of the bidirectional relationship between HF and diabetes. People with HF and diabetes are at higher risk for all cause and CV mortality, while people with severe HF are at higher risk for developing diabetes. Of the existing classes of anti-diabetes drugs, metformin, SGLT-2 inhibitors, and GLP-1 agonists have been associated with decreased risk of heart failure. 

  • SGLT-2 inhibitors are considered first line therapy under the new ESC guidelines based on positive data from EMPA-REG, CANVAS, CREDENCE, DECLARE, and most recently, DAPA-HF

  • GLP-1 agonists are a “promising group,” but ultimately were classified as neutral regarding HF. Some class members, like liraglutide, have been linked to non-significant increases in rates of HF hospitalizations (HHF) and composite death or rehospitalization. A recent meta-analysis, however, has shown a a 9% reduction in HHF with GLP-1. ESC guidelines classify them as neutral regarding HHF, and say they may be considered for diabetes patients with HF. We note that data recently presented at the 2019 World Congress on Acute Heart Failure show that, in the HARMONY CVOT for GSK’s now-discontinued Tanzeum, albiglutide gave a 29% relative risk reduction on heart failure hospitalization vs. placebo (HR=0.71, 95% CI: 0.53-0.94, p<0.02). Could there be an untold story with GLP-1 with heart failure? What a loss from our view that GSK opted to drop the drug – we believe it could’ve been a major opportunity for Wal-Mart or a low-cost retailer to provide some public health help (it would not be as popular as the other GLP-1s, presumably, due to less weight loss and overall glycemic potency associated with it overall – the cardioprotection would seem quite valuable however).

3. KOLs React to ESC 2019 Guidelines: Mixed Opinions on First-Line Use of SGLT-2/GLP-1

Multiple key opinions leaders provided their thoughts on the new ESC guidelines, focusing on the recommendation to use SGLT-2/GLP-1 as first line therapy over metformin in patients with established CVD or at high CV risk. See below!

  • “I think the revised ESC guidelines are forward-thinking but may push the envelope a bit too much.  

First, there is only inferential evidence that either an SGLT2 inhibitor (i) or a GLP-1 receptor agonist (RA) will have a CV benefit when used as monotherapy before metformin - mainly from the statistical lack of heterogeneity when CV outcomes are assessed in participants taking vs. not taking metformin.  Remember, however, that across all recent CVOTs, about 70-75% of patients were already on that "foundation therapy." So, it could be stated that the trials have proved that the newer agents reduce MACE mainly when used in combination with metformin. Of course, it is true that some 25-30% were NOT on metformin, but I suspect the vast majority of these either had CKD rendering metformin relatively contraindicated or could not tolerate the drug in the past. So, perhaps it’s more precise to say that the trials prove that the newer agents reduce MACE mainly when used in a population where the vast majority of patients who could take metformin did.  The new ESC guidelines shake that up a little and recommend going straight to either an SGLT2i or a GLP-1RA before metformin  A specific study would need to be carried out before we recommend these very expensive agents be used so early.  The ESC may actually be correct – but this recommendation is just not yet evidence based.

Second, the ESC guidelines seem to suggest using medications even when the diabetes is already well-controlled.  This is in contrast to the ADA-EASD consensus report from 2018.  Most of the CVOT’s have had an HbA1c <7.0% as an exclusion criterion, so evidence for benefit when glycemia is already controlled is not that robust.  Our group reported that empagliflozin’s CV benefits in EMPA-REG OUTCOME had very little to do with HbA1c control, so I actually suspect the ESC may again be correct here. But, I prefer solid evidence, and for that, you’d need a separate trial.  Moreover, I don’t understand their recommendation to use metformin in a drug-naïve patient without ASCVD even if the HbA1c is already controlled. There is little evidence for any benefit in this circumstance.

Third, the new guidelines lump in ASCVD patients with those at ‘high or very high CV risk’. While there is some evidence with SGLT2i’s that HF is prevented and CKD progression is slowed in both those with ASCVD and those with just multiple risk factors, the same is not true for MACE, where the non-ASCVD patients appear to have little benefit.  With GLP-1 RA’s most of the evidence suggests that those without ASCVD don’t benefit. REWIND may be an exception.  I would have preferred a more cautious statement about using these drugs in those without overt CVD.

I have a few other more minor observations. The ESC is now quick to advise combination therapy involving an SGLT2i plus a GLP-1RA.    It’s also interesting to see that the ESC has placed TZDs much higher than either SU's or insulin (unless HF is present) – probably on account of the findings from the PROactive and IRIS trials. . Finally, the ESC algorithm figure does not give enough guidance to which agent (SGLT2i or GLP-1RA) should be used in which type of ASVCD patient. Very clearly, if HF or mild-moderate CKD is present, the SGLT2i’s should be the go-to drug. 

So, in summary, the ESC guidelines are overall nicely done but may be overly aggressive in parts and at points not evidence-based enough. However, I suspect they will eventually be proven right!” – Dr. Silvio Inzucchi (Yale University)

  • “It is certainly noteworthy that the new guidelines have made this change. In the past guidelines have said that SGLT2 inhibitors or GLP1RAs should be add-ons to metformin in order to be consistent with the design of the trials. Importantly, however, subgroup analyses from all of the trials have shown similar benefits in the smaller groups of patients not on background metformin. Since the cardiologist may not continue to follow the patient on a long term basis, omitting the requirement that patients already be on metformin will hopefully get more patients on evidence based therapies. In addition, industry has been reluctant to do head-to-head trials vs. metformin as initial therapy fearing that even positive results might not supplant metformin’s role as the foundational therapy. Hopefully these new guidelines may encourage proper comparative trials to now be initiated.” – Dr. Lawrence Leiter (University of Toronto)

  • This is a bold move by the ESC and I for one, commend them for taking this step. Event reduction should be a priority when selecting therapy for a given patient - be it pharmacological or nonpharmacologic. After therapies to lower risk are implemented, then add whatever else is needed to achieve glycemic control.  I hope that this will lead the way for other guidelines to stop the “love affair” of Metformin and make the outcome- reducing therapies more accessible and achievable.” – Dr. Alice Cheng (University of Toronto)

  • "The ESC group has clearly thought carefully about their recommendations. Though I think they are acceptable, they do go a bit beyond the available data with best intentions. The vast majority of patients in the CVOTs have been on metformin at baseline. For example, DECLARE and REWIND were the trials with the widest enrollment criteria and both had >80% of patients treated with metformin at baseline. That said, I know of no subgroup analyses that demonstrate an interaction for CV effects based on baseline metformin status. Metformin does have suggestive evidence for benefit for CV events. Thus, prescribing a GLP-1RA or SGLT2i as first line therapy in patients with type 2 diabetes at high CV risk is practicing in an evidence-free zone. It is not an unreasonable recommendation; it is just not evidence-based.-- Dr. John Buse (UNC Chapel Hill) 
  • This recommendation was a long time coming and is faithful to the evidence. There is little to no evidence supporting the use of metformin in mitigating cardiovascular risk in patients with diabetes. On the other hand, there is clinically compelling and statistically persuasive evidence for CV risk reduction with SGLT2 inhibitors and GLP1-RA in T2DM patients with underlying ASCVD. The fact that metformin continues to enjoy status as first-line therapy in current guidelines is an example of old habits die hard! While metformin is affordable and safe and physicians are facile with its use, the evidentiary basis for its recommendations rests on a shaky foundation. In EMPA-REG OUTCOME trial, about a quarter of the patients were not on metformin at baseline, and this subgroup experienced a greater risk reduction in the primary endpoint (28% vs 8%) or CV death (54% vs 29%) than the subgroup on metformin at baseline. These data support the use of empagliflozin as monotherapy in patients with established ASCVD. I will not be surprised to see other guidelines following the lead of ESC recommendations.” – Dr. Sanjay Kaul (Cedars-Sinai Medical Center)

  • "I think that recommening SGLT-2i and GLP1 RA as first line therapy in patients with T2D and ASCVD (or high CV risk) is a huge step in the right direction – and is based on consistent evidence from multiple large clinical trials. It’s great to see that the guidelines continue to emphasize the selection of agents with proven cardiovascular benefit for high-risk patients with type 2 diabetes. Since cardiovascular disease remains the main cause or morbidity and mortality in this patient group, prevention of these common and morbid complications should be the key goal of management." -- Dr. Mikhail Kosiborod
  • "While acknowledging the exciting results of the cardiovascular outcome studies with SGLT2i and GLP-1RAs in the appropriate population, it is worth noting that the majority of people with diabetes studied in these trials were already on metformin-our database for robust cardioprotection with these 2 classes of drugs as first-line therapy alone in the absence of metformin use is much less scientifically compelling" – Dr. Daniel Drucker (University of Toronto)

  • “ESC is composed mainly of cardiologists and subsequently has a strong cardiology focus. From a cardiology perspective, I understand why ESC made SGLT-2 inhibitors and GLP-1 receptor agonists first-line ahead of metformin. Cardiologists focus more on heart disease and heart failure. SGLT-2 inhibitors and GLP-1 receptor agonists have shown proven benefit for people with those conditions.  However, the focus of endocrinologists is hyperglycemia. Controlling A1C and hyperglycemia remains their number one priority. Many people will end up being prescribed both metformin and a SGLT-2 inhibitor or a GLP-1 receptor agonist. It may not really matter which one is first in that situation

It is certainly possible that ADA/EASD will make SGLT-2 inhibitors or GLP-1 agonists first-line over metformin in the future. However, I don't expect such a change simply because of what ESC has done. Metformin is also much less expensive than SGLT-2 inhibitors or GLP-1 agonists in the USA at this time which will also help to keep it first-line for the present.” – Dr. Charles Alexander

  • “It’s about time that a guideline skipped metformin as first line therapy. The pathetically small metformin cohort in UKPDS does not compare with the current generation of CVOTs.” – Dr. Jay Skyler (University of Miami)

  • “The ESC has provided us with a remarkable and thoughtful document spanning a variety of aspects of the relationship between diabetes and CVD. Among the highlights of this very practical set of recommendations are:

    • In CV risk assessment, the recommendation that a resting ECG be done as an aspect of regular screening.

    • In glycemic control, the recommendation that a goal A1c<7% not only be part of the prevention of microvascular complications, but also should be considered for prevention of macrovascular complications – particularly in persons with long life expectancy.

    • The note that a goal A1c<6.5% “may be suggested on a personalized basis.

    • The GTT is suggested as being useful in diabetes diagnosis.

    • The recommendation that SMBG and/or continuous glucose monitoring “be considered to facilitate optimal glycemic control.”

    • Useful clarification of blood pressure treatment targets and approaches, and recognition of potential roles of self-monitoring of blood pressure.

    • The assertion, along the lines of those from other groups, that SLP-1RA’s and SGLT2i’s should be recommended “in patients with T2DM with prevalent CVD or very high/high CV risk.”

    • The ranking of sulfonylureas after DPP4i, SLP-1RA, SGLTT2i, and TZD because of their association with hypoglycemia – although recognizing that there was no evidence of worse CV outcome with a SU than with a DPP4i in CAROLINA.

    • The assertion that “First-line treatment of DM in HF should include metformin and SGLT2 inhibitors; conversely, saxagliptin, pioglitazone, and rosiglitazone are not recommended;” interestingly, the document explicitly note sitagliptin and linagliptin not having adverse effect on HF. SGLT2i are recommended to reduce HF risk, but not specifically for treatment of HF.

    • For CKD, SGLT2i are “recommended if eGFR is 30 to <90 mL/min/1.73 m2,” and the GLP-1 liraglutide and semaglutide are suggested for GFR>30.” – Dr. Zachary Bloomgarden (New York University)

4. Entresto Narrowly Misses Superiority in PARAGON-HF Trial of HFpEF Patients

The full readout of Novartis’ PARAGON-HF trial (n=4,822) revealed that Entresto (sacubitril/valsartan) just missed its primary endpoint of composite heart failure hospitalization and CV death in HFpEF patients. Full results were simultaneously published in NEJM – see here. Topline results were released last month. Treatment with sacubitril/valsartan conferred a just barely non-significant 13% reduction in the primary endpoint (RR: 0.87; 95% CI: 0.75-1.01; p=0.059). However, benefit was observed in the secondary endpoints (quality of life, change in NYHA classification, and renal function) for women and patients with lower ejection fractions (≤ 57%)—of course, these are exploratory endpoints given the miss of superiority on the primary outcome. Overall safety and tolerability was similar to that presented in the PARADIGM-HF trial (Entresto in HFrEF), but with higher incidence of hypoglycemia and less elevation in serum creatinine and potassium levels. Thus, results are mixed, and some patients may benefit more from treatment with sacubitril/valsartan vs. valsartan alone. Positive results in PARAGON-HF would have represented the first ever for a potential treatment of HFpEF. We note that SGLT-2 inhibitors are also under investigation specifically in this population of people with HFpEF (EMPEROR-Preserved for Lilly/BI’s Jardiance, DELIVER for AZ’s Farxiga), contributing to a recent push to help fill this vast unmet need.

  • Introduction and Research Question: Dr. Scott Solomon began the session with an overview of the investigation of sacubitril/valsartan in different forms of heart failure. Previously in the PARADIGM-HF study, this neprilysin inhibitor and angiotensin receptor blocker combination therapy reduced morbidity and mortality in patients with heart failure and reduced ejection fraction (HFrEF), gaining a class I guideline recommendation in the process. PARAGON-HF was designed to determine if sacubitril/valsartan can improve outcomes similarly in HFpEF patients.

  • Background, Design, and Outcomes: Over 64 million people worldwide have heart failure, and this number is only set to increase as the population ages. Currently, valsartan is the only therapy backed by evidence to treat HFrEF patients. PARAGON-HF was designed to create another evidence-based treatment option for patients with HFpEF and meet the large, unmet need for treatment in that population. PARAGON-HF is a randomized, double blind, active comparator trial of sacubitril/valsartan (97 mg/103 mg twice daily) against valsartan (160 mg twice daily). The primary outcome was composite total HF hospitalizations and CV death, while secondary outcomes included improvement in NYHA class at eight months, changes in KCCQ clinical sum score, time to first occurrence of worsening renal failure, and time to all-cause mortality. All enrolled patients were at least 50 years old, had an EF ≥ 45%, had symptoms of HF, had structural heart disease, had elevation in natriuretic peptides, did not have decompensated HF, and did not have any other reason for their symptoms outside of HF.

  • Outcomes: Sacubitril/valsartan did not meet its primary endpoint of reducing CV death and total HF hospitalizations, narrowly missing significance at a RR of 0.87 (95% CI: 0.75-1.01; p=0.059). The Kaplan Meier-curve for the composite primary outcome can be seen below. Composite reduction was mainly driven by a reduction in first and recurrent HF hospitalizations (690 hospitalizations for heart failure and 204 deaths from cardiovascular causes). Interestingly, the primary endpoint was found to be much more significant amongst women (RR=0.73, 95% CI: 0.59-0.90) than men (RR=1.03, 95% CI: 0.85-1.25). 

    • Dr. Solomon emphasized the possible heterogeneity of treatment effect by highlighting that benefit was observed in some secondary outcomes such as quality of life, change in NYHA classification, and renal function for women and patients with lower ejection fractions (≤ 57%), although these observations are exploratory since primary outcomes were not met. In the sacubitril/valsartan group, 15.0% of patients had an improvement in NYHA class vs. 12.6% in the valsartan group. 

    • All in all, Dr. Solomon noted that the findings suggest that some patients with HFpEF, specifically those with ejection fractions below normal but not fully reduced, may benefit from sacubitril/valsartan. 

  • Safety: Overall safety and tolerability of sacubitril/valsartan were comparable to previously reported findings in HFrEF patients in the PARADIGM-HF Trial. Patients on sacubitril/valsartan had higher rates of hypertension (380 vs. 257 events in valsartan; p<0.001)  but lower rates of increase in creatinine and potassium levels. 25% of patients in the sacubitril/valsartan group and 27% of the valsartan group discontinued the trial drug for reasons other than death. 15% of patients on sacubitril/valsartan discontinued the drug because of an adverse event, while 16% did so on valsartan. 

  • Independent Commentary: Dr. Stuart Connolly conducted the discussant review on the PARAGON-HF data presented. He began his review with praise for the study’s design and conduction, highlighting that HFpEF patients were confirmed to have the condition by multiple measures (symptoms, majority recent hospitalizations, abnormal structure on Echocardiograms). This was further emphasized with a very low 0.2% attrition rate. He went on to state that the overall treatment effect was modest and pointed out an almost 13% significant reduction in hospitalizations and CV death was found against an active control arm despite valsartan not yet being approved to treat HFpEF. One of the most striking aspects of PARAGON-HF was its subgroup analyses, and Dr. Connolly pointed out that this requires caution, seeing as these analyses are exploratory in nature. Validation of sub-group interaction usually requires pre-specification, statistical significance, biological plausibility, and observations in prior studies. With this, it’s possible that EF abnormalities were present in the HFpEF group, as slight EF abnormalities have been seen patients with structural abnormalities on Echocardiograms in former studies like AZ’s CHARM trial and the NIH’s TOPCAT study. More focus should have been placed on subgroup analyses, and despite Novartis presenting promising outcomes in women and patients with lower HFpEF, this doesn’t outweigh the overall modest effect driven by Entresto.

5. New Risk Score Developed for Heart Failure in Diabetes: Easy-to-Use Tool with Potential to Identify Patients that Derive Greatest Benefit from Treatment

Dr. David Berg detailed the newly developed TIMI Risk Score for Heart Failure in Diabetes for use as a clinical tool in predicting HHF in people with diabetes. Notably, applying this tool to data from DECLARE highlighted useful risk stratification and identification of patients who may benefit most from dapagliflozin treatment. Risk stratification and patient identification remain an important area of research concerning how to best implement highly efficacious (yet expensive) diabetes therapies such as SGLT-2 inhibitors – as a result, we’re glad to see an emphasis in developing new tools to better address this area. See Dr. Berg’s Twitter thread explaining his work, along with our summary and takeaways below. This work was also simultaneously published in Circulation.

  • The tool takes in five clinical variables that are routinely assessed: (i) prior heart failure (two points); (ii) atrial fibrillation (one point); (iii) coronary artery disease (one point); (iv) eGFR below 60 (one point); and two points for an UACR >300 or one point for UACR between 30-300. Adding the point values for each of these variables together for each patient gives the patient their individual Risk Score.

  • When applied to the patient population from DECLARE, this new risk score accurately identified patients that may benefit the most from dapagliflozin treatment. Relative risk reductions with dapagliflozin treatment vs. placebo were consistent across each risk score group (highlighting dapagliflozin’s consistent effect across the risk spectrum), but absolute risk reductions were greatest in the highest risk cohort (see graphic below). In practice, we see great potential for this tool to allow clinicians to quickly identify which of their patients are at highest risk and would derive the greatest benefit from SGLT inhibitor treatment. We can’t help but think about how nicely this tool dovetails with the goals set out by new ESC diabetes guidelines that recommend first-line use for patients at high CV risk – could this tool help to better pinpoint those high risk patients that would benefit most? Can similar tools be developed for MACE risk, along with HF? It seems like it!

A standing-room only audience packed this Medscape LIVE! interactive session on GLP-1 receptor agonists sponsored by Novo Nordisk. We’re always curious to see how drugs primarily developed to treat diabetes are being translated to the broader audience, and from this session, it’s clear that there’s still a very strong (and much needed) effort to teach GLP-1 basics to the cardiology community (as well as the endocrinology and PCP communities) – many questions from cardiologists during Q&A  still focused on foundational ideas, i.e. “how do GLP-1s compare to DPP-4 inhibitors?” (We found that one quite shocking.) On top of that, this session served to highlight the ongoing movement to emphasize comprehensive CV risk reduction from GLP-1s that goes beyond glycemic control. Highlights from the panel included (i) Dr. Juris Meier on incretins and glucose homeostasis; (ii) Dr. Mansoor Husain on GLP-1 and the CV system; and (iii) Dr. Agostino Consoli on oral GLP-1s in CV risk management.


Dr. Meier touted GLP-1 as almost an “ideal candidate for the treatment of diabetes,” due to its ability to (i) act as a satiety factor; (ii) slow entry of nutrients into the circulation by delaying gastric emptying; and (iii) control postprandial glucose homeostasis by increasing insulin secretion and reducing glucagon secretion. Although Dr. Meier covered the glucose-lowering and substantial weight loss effects seen in patients with type 2 treated with GLP-1s, he was also sure to emphasize that the compound has a diverse array of targets in the body, including the CV system and central nervous system. In particular, semaglutide (Novo Nordisk’s Ozempic) was underscored as being more efficacious than all other comparators, including dulaglutide (Lilly’s Trulicity) as seen in SUSTAIN 7. Ozempic is more expensive in many parts of the US, which prevent many patients from getting access to it.  


Quelling skepticism surrounding GLP-1s due to increased heart rate (HR) was distinctly a focus of Dr. Hussain’s talk. We haven’t seen as much attention on this particular risk signal in purely diabetes-oriented settings and felt the distinction here at ESC and were slightly surprised by it. About ~1500 patients in the LEADER trial for liraglutide experienced a ≥10 bpm increase in heart rate, but crucially, a consistent MACE and hospitalization for heart failure (HHF) benefit was seen in both patients with heart rate increases <10 bpm and ≥10 bpm, suggesting that the HR change is not only well-tolerated but does not seem to affect MACE or HHF.

  • Dr. Hussain also highlighted that GLP-1s clearly have mechanisms beyond blood systems and metabolic lowering, detailing both human and animal studies. Up-and-coming areas of interest included sodium secretion, diuresis, anti-atherosclerotic and anti-inflammatory effects, release of atrial natriuretic peptide and subsequent loss of visceral adiposity, reduction of carotid intima medial thickness, and diminished thrombosis (a benefit distinct from SGLT-2 inhibitors).

  • Dr. Consoli rounded out the session by detailing the mechanism and benefits of oral semaglutide, as seen in PIONEER 6. Apart from extolling oral semaglutide’s clear benefit to patient adherence over injectables, a good portion of his allotted time was dedicated to explaining how SNAC-mediation allows for (i) changes in local stomach pH to reduce degradation; (ii) an increase in the monomeric form of semaglutide that can be absorbed; and (iii) changes to gastrointestinal epithelium fluidity that increase absorption. Dr. Consoli noted that a CVOT for oral semaglutide (SOUL) has been initiated, and the next challenge is proving that oral semaglutide demonstrates not only non-inferiority vs. placebo, but superiority. The superiority-powered trial (n=9,642) enrolling people with type 2 and either CHD, CKD, PAD, or cerebrovascular disease is set to complete in July 2024, and if positive will likely support CV indications for both Ozempic and oral semaglutide (assuming that FDA does not grant CV indications for both of these therapies based on the current sNDA submission using data from SUSTAIN 6 and PIONEER 6).

Dr. Subodh Verma presented compelling pre-clinical data showing that SGLT-2 inhibitor empagliflozin significantly reduces pulmonary arterial hypertension-related mortality and progression in rats without diabetes. As a reminder, the EMPA-REG outcome trial previously demonstrated that empagliflozin reduces hospitalization for heart failure by 35% in individuals with type 2 diabetes (HR = 0.65; 95% CI = 0.50-0.85; p=0.002), although the particular mechanism of action is still unknown. Notably, rats in Dr. Verma’s study did not have diabetes, and we’ve definitely seen an excitement toward SGLTs from the cardiology community here at ESC in patients without diabetes (look no further than landmark DAPA-HF results just presented). The purpose of the study was to investigate empagliflozin’s effect on pulmonary arterial hypertension (PAH) – a condition with few current therapeutic targets. PAH was induced in male rats (n=66) using intraperitoneal injection of monocrotaline (MCT) at 60 mg/kg.

  • Rats treated with empagliflozin (n=8) had a significantly improved median survival rate of 33 days vs. 24 days (p<0.05) in vehicle-treated rats (n=8). Empagliflozin (10 mg/kg) was administered daily by oral gavage over the course of 28 days, and rats were monitored for up to 45 days following MCT-injection at day 0.

  • Empagliflozin delayed the progression of MCT-induced pulmonary hypertension. Right ventricular systolic pressure and pulmonary artery pressure were both significantly reduced in empagliflozin-treated rats, while pulmonary acceleration time was significantly increased compared to vehicle-treated rats.

  • Morphologically, empagliflozin also inhibited MCT-related arterial remodeling. Rats treated with empagliflozin had significantly less medial wall thickening (50.8 ± 2.2 vs. 44.7 ± 1.1 mm) and pulmonary artery muscularization (53.2 ± 1.3 vs. 43.6 ± 2.1 mm) than vehicle controls.

Overall, Dr. Verma encouraged greater investigation into treatment of PAH using empagliflozin. In particular, teasing out whether benefits are primarily cardiac, pulmonary, or hemodynamic seemed of the greatest interest. Dr. Verma was also sure to acknowledge the limitations of the study including that (i) MCT-induced PAH has pathophysiological differences to human PAH; (ii) models of hypoxia or PA banding are needed to reinforce results; and (iii) a dose responsive curve to empagliflozin was not yet performed.

  • Looking at the greater landscape, MannKind currently has a partnership with United Therapeutics to develop Trepostinil Technosphere technology for PAH, and Arena’s potent IP receptor-agonist ralinepag is in phase 3 testing. Gilead chose not to pursue phase 3 PAH development for its ASK-1 inhibitor selonsertib in favor of NASH, which recently failed to meet primary endpoints in STELLAR 3 and STELLAR 4.


--by Ursula Biba, Rhea Teng, Martin Kurian, and Kelly Close