Hello from New Orleans, where the first two days of ACC 2019 took place this weekend! Below, you’ll find our top nine highlights from the start of the meeting; tomorrow, we’ll be back with exciting analyses from DECLARE (consistent heart failure benefit across ejection fractions; significant MACE benefit in those with prior MI) plus coverage of the exhibit hall. Read on!

Top 9 Highlights

1. Apple Heart Study Results Bring Excitement and Skepticism; 0.5% of >400,000 Participants Notified of Potential Atrial Fibrillation; 16% False Positive Rate; >20,000 PWD in Study

To a packed-to-capacity Main Hall in the first session of ACC 2019, Stanford’s Drs. Mintu Turakhia and Marco Perez presented highly-anticipated results from the Apple Heart Study, a pragmatic trial designed to validate the use of Apple Watch data to detect atrial fibrillation (AF). The trial impressively enrolled 419,297 people with an Apple Watch in just over eight months (wow!), collected its last piece of data on February 25 of this year, and read out less than a full month later. Beyond the results from the study – more on this below – we think this is a tremendous step toward in virtual studies, and the remarkable speed, efficiency, size, and diversity that can be achieved through this methodology when compared to standard trial designs is undeniably exciting. How can diabetes studies similarly leverage some the successful aspects of this approach?

Assuaging some fears over a potential over-burdening associated with possible AF notifications from the Apple Watch, results indicated that only 0.5% of participants (n=2,161) were notified at any point of potential AF:

Moving forward, the big question will be how cardiologists will interpret and handle a potential flood of data from Apple Watches, which now have the ability to take an ECG and notify wearers of irregular heart rhythm (the Apple Heart Study importantly did not use the latest Apple Watch with the ECG feature). How will cardiologists (and PCPs) react to a patient coming in with a potential AF signal, especially considering the likelihood of a false positive (16% of notifications in the study did not yield confirmation of an actual AF diagnosis, although these did help in identifying other potential arrhythmias). Will this become a drain on the healthcare system, with many otherwise healthy individuals soaking up cardiologists’ already-scarce time? On this point, it is notable that only 0.16% of those those <40 years old were notified during the study.

• The study enrolled over 20,000 people with diabetes (~5% of the total cohort), and 12% of the “notification” group (n=2,161) was comprised of people with diabetes (n=255). When compared to the overall population, those with diabetes were at a greater risk of being notified about potential atrial fibrillation (0.52% vs. 1.25%). This enrichment of people with diabetes in the notification group is unsurprising, as the trend falls in line with past research that has established diabetes as a risk factor for AF. Surveys of hospitalized patients with hypertension have shown that AF occurs in 15% of people with diabetes and 10% of people without diabetes (in a higher-risk group).

  • Over 160,000 people with obesity were enrolled in the study (38% of total trial population), with 46% of the notification group comprised of those with obesity (n=984).

• Apple and partner J&J recently announced their next step in leveraging the Apple Watch in personal health through the HEARTLINE study. HEARTLINE will build upon results from the Apple Heart Study but will importantly adopt a randomized, controlled design – people will either be assigned to the group with the Apple Watch + J&J’s heart health app, or to nothing at all. The study will only include people older than 65 and is set to begin later this year. Notably, J&J is keenly interested in better detection of AF, seeing as it co-markets blood thinner Xarelto (rivaroxaban) and is currently developing additional therapies for stroke.

• For additional analysis and commentary on the study, we point to two excellent articles in STAT by Matthew Herper and Larry Husten. Of note, Dr. Steven Nissen is quoted in Mr. Herper’s piece, and sounds skeptical of the results and impact of the study. Dr. Nissen emphasized that the data are still “unpublished, unverified, unreviewed, and not tested by peers,” and also worries that this avalanche of data could be an enormous cost to the system: “How much money is it going to cost the health care system to follow up on thousands of patients who have an incidental atrial fibrillation?” 

2. CANVAS Analysis Suggests Similar Heart Failure Benefits with Invokana Across Different Levels of Ejection Fraction

Dr. Kenneth Mahaffey presented a CANVAS analysis showing consistent effects of J&J’s Invokana (canagliflozin) on heart failure (HF) events (death or hospitalization) across both reduced and preserved ejection fraction. While optimistic, he cautioned that these are hypothesis-generating data only. As a reminder, the CANVAS program found a 33% relative risk reduction on hospitalization for heart failure (HR=0.67, 95% CI: 0.52-0.87) and a 30% risk reduction on hospitalization for or fatal HF (HR=0.70, 95% CI: 0.55-0.89) with canagliflozin vs. placebo, in a population with ~two-thirds prevalence of cardiovascular disease at baseline. And while there was a benefit in patients with and without a history of HF at baseline, it was initially unclear (and, to some degree, still is) whether the heart failure benefit seen with the SGLT-2 applies to those with HF with reduced ejection fraction (HFrEF), HF with preserved ejection fraction (HFpEF), or both (note that those with NYHA Class IV HF were excluded). This question is key, as there are currently no approved therapies for HFpEF, and this is the type of heart failure thought to co-occur with obesity and more frequently impact patients with diabetes – making the potential of SGLT-2 inhibitors particularly promising. As presented by Dr. Mahaffey, among CANVAS’ 10,142 participants, 276 had a fatal or hospitalized heart failure event; 101 had a first HFpEF event, 122 a first HFrEF event, and 61 were unknown (HFuEF). Interestingly, 8 participants had multiple types of HF events, and 52 experienced >1 HF events. This analysis includes the first HF event of each type in a given patient (summing to 284 events). Canagliflozin reduced risk across all types of events, with a slight trend toward greater risk reduction in HFrEF events – though the most impressive point estimate was in HFuEF, reflecting the hypothesis-generating nature of these findings. Ultimately, while the confidence intervals (see below) overlap, Dr. Mahaffey emphasized the need for additional data to confirm canagliflozin’s efficacy in both types of heart failure (pointing to presentation of CREDENCE, J&J’s renal outcomes trial for Invokana, at WCN 2019 next month).

• Because CANVAS protocol did not stipulate EF measurement either at baseline or during HF events, a blinded Adjudication Committee member (Dr. Gemma Figtree) conducted a retrospective, secondary review of medical records to assign HF events post-randomization. For the purpose of this analysis, HFpEF was defined as no prior report of reduced EF and with documentation of an EF ≥50% during the HF admission. HFrEF was defined as either documentation of an EF <50% during admission or a prior report of reduced EF without evidence of recovery. Any other HF events were defined as unknown (HFuEF).

3. Observational Study Finds Consistent Heart Failure and Mortality Benefits in Patients with Heart Failure with Preserved or Reduced Ejection Fraction

Dr. Mikhail Kosiborod presented results from an observational, AZ-sponsored study indicating that the benefits of SGLT-2 inhibitors on heart failure and all-cause death may extend across the range of baseline ejection fraction. The study propensity matched 10,614 patients (5,307 on SGLT-2 inhibitors and 5,307 on another glucose lowering drug) from Maccabi database in Israel, who had documented left ventricular ejection fraction (LVEF) data at baseline. LVEF was categorized as either preserved (≥50%) or reduced (<50%). Notably, baseline characteristics were well balanced between the two groups (<10% standardized difference), except socioeconomic status, which was slightly higher in the SGLT-2 inhibitor group – raising the question as to how much of an effect this imbalance may have had in driving overall treatment benefit, though it seems unlikely to us that this would have impacted any difference or similarity across ranges of LVEF. Patients in the SGLT-2 inhibitor group were on either Jardiance (empagliflozin) or Farxiga (dapagliflozin), and median follow-up time in the study was ~1.5 years for both groups. On the endpoints of hospitalization for heart failure (HHF), all cause death (ACD), or a combination of the two, SGLT-2 inhibitor use was associated with a consistent, statistically significant benefit (see figure below). Importantly, this benefit did not appear to meaningfully change depending on EF status, either reduced or preserved. Effects were also consistent when EF was dichotomized using the 40% cut-off (instead of 50%), when restricting analyses to those with a history of HF, and when using EF measurements within 2 or 4 years of the index date (data not shown).

• As evidence has emerged on the impressive benefit that SGLT-2 inhibition has on heart failure outcomes, manufacturers have heavily invested in further studying the potential link in larger studies: AZ has launched DELIVER and Dapa-HF to investigate dapagliflozin in HFpEF and HFrEF, Lilly/BI have two trials underway for Jardiance in HFpEF (EMPEROR HF-Preserved) and HFrEF (EMPEROR HF-Reduced), and Sanofi/Lexicon are studying sotagliflozin in worsening HF (SOLOIST-WHF). J&J’s CREDENCE renal outcomes study has secondary endpoints examining heart failure. We’re looking forward to readouts from these studies in the coming years (and CREDENCE next month) to provide a more definitive answer on HFrEF and HFpEF impact with this class of agents.

• We’ll get even more insight into the relationship between SGLT-2 inhibitor treatment and outcomes in HFrEF and HFpEF on Monday, when AZ is set to present a new prespecified analysis of DECLARE data on this subject.

4. ACC/AHA Cholesterol Guidelines: Dr. Sperling Emphasizes Diabetes as a Key CV Risk Factor, Calling for Comprehensive Risk Reduction; Emerging Importance of Triglycerides?

Dr. Laurence Sperling presented the diabetes-specific components of ACC/AHA’s new guidelines on lipid lowering therapy, highlighting that many people with diabetes should automatically be prescribed a moderate-intensity statin and underscoring the importance of comprehensive, event-reduction-focused treatment. A top 10 take-home message from the guideline update is that, in those 40-75 years old with diabetes and LDL-C ≥70 mg/dl, HCPs should start a moderate-intensity statin without calculating 10-year ASCVD risk. Dr. Sperling clarified that this doesn’t mean 10-year risk should never be calculated – rather, it means these patients are at a baseline risk level that calls for statin use. When 10-year risk is calculated, patients with diabetes at higher risk (especially those with multiple risk factors or those 50-75 years old), it’s reasonable to use a high-intensity statin to achieve ≥50% LDL-C risk reduction. Additionally, it’s critical to note that while diabetes itself is a risk factor, there are also diabetes-specific risk enhancers: longer diabetes duration (≥10 years for type 2 and ≥20 years for type 1), UACR ≥30 mcg/mg, eGFR <60 ml/min/1.73 m², retinopathy, neuropathy, and ankle-brachial index <0.9 – these can all contribute to CV risk and mean that more aggressive lipid-lowering therapy is appropriate. Dr. Sperling also emphasized that, within the spectrum of secondary prevention patients, those with diabetes comprise a particularly high-risk cohort that necessitates aggressive treatment. For patients with multiple major ASCVD events or one event and multiple high-risk conditions (“very high risk” – a group newly formed in this guideline edition), maximally-tolerated statin therapy is recommended with the possibility of adding ezetimibe and then a PCSK9 inhibitor if LDL-C remains above 70 mg/dl; for other secondary prevention patients (“stable ASCVD”), the guidelines recommend high- or moderate-intensity statin therapy (see the full algorithm on page 7).

• When it comes to older and younger patients, there is more room for clinical judgment but statins can still be beneficial. In those 20-39 years old, Dr. Sperling noted that statins are a reasonable treatment addition for those with a longer duration of diabetes, microvascular disease, or an ABI <0.9. On the other end of the spectrum, statins can also be used in those over 75 years old following a discussion between the HCP and patient. Above all, Dr. Sperling cautioned against “de-risking” individual patients with diabetes who are at intermediate risk (for example, with a coronary artery calcium score of 0) – diabetes in and of itself is a risk factor that needs to be taken seriously. Additionally, he minimized the notion that “statins cause diabetes,” explaining that statins can “nudge people at-risk over the threshold” but this shouldn’t be a reason to withhold or discontinue statin therapy, as the net clinical benefit is still favorable – he also highlighted this conversation as an opportunity to discuss diabetes prevention with patients, which we thought was very valuable for cardiologists to hear.  

• Dr. Sperling also gave a nod to triglyceride lowering, explaining that hypertriglyceridemia is common in those with type 2/metabolic syndrome. He recommended lifestyle changes as the primary treatment strategy for high triglycerides (with an emphasis on secondary factors like hyperglycemia), with possible use of omega-3’s or fibrates if persistently severe (>500 mg/dl). We anticipate increased attention to the issue of triglycerides in coming years, in the wake of positive results from the REDUCE-IT CVOT for Amarin’s Vascepa (icosapent ethyl, a highly-purified component of omega-3’s). In REDUCE-IT, a highly-significant 25% relative risk reduction (HR=0.75, 95% CI: 0.68-0.83) on the primary MACE endpoint (CV death, MI, coronary revascularization, and unstable angina) was ostensibly driven, at least in part, by an impressive reduction in triglycerides (median 18% reduction, or -39 mg/dl from a baseline of 216 mg/dl, at one year).

5. Diabetes and CVD Roadmap from WHF & IDF Aims to Improve CV Outcomes through Patient-Centricity, Inter-Professional Collaboration

At a preview of the World Heart Federation’s Roadmap for cardiovascular disease prevention in people with diabetes, Dr. Laurence Sperling (Director of the Emory Heart Disease Prevention  Center, Professor of Global Health, Hubert Department of Global Health, Rollins School of Public Health at Emory University), outlined progress to date and moderated a panel comprised of Ms. Emily Fitts (Senior Manager, Advocacy and Policy, The diaTribe Foundation), Dr. Kornelia Kotseva (Senior Clinical Research Fellow, National Heart & Lung Institute), and Mr. Richard Wood (CEO, dQ&A). This “Roadmap” – a joint effort with IDF – is one of nine in the WHF’s “Roadmap family” aimed at driving sustainable, policy-based action to reduce the burden of non-communicable diseases, spanning from hypertension and heart failure to Chagas disease and rheumatic heart disease. To date, the writing committee (comprised of writers from six continents) has met to brainstorm and map out key priorities, roadblocks and solutions, and unique features focused on implementation and patient centricity. On the last point, Dr. Sperling pointed to the partnership of WHF with The diaTribe Foundation, which he emphasized as an important step in making sure patient perspectives are represented. The group hopes to have a document published by May 2019 and is aiming for a “side event” at the World Health Assembly in Geneva that month. Official launch should follow in August 2019 with a final publication in Global Heart, and we’re anticipating events at ESC 2019 and IDF 2019. Of course, implementation is key, and WHF CEO Jean-Luc Eiselé spoke to this consideration: WHF is planning to focus on local implementation and convene national roundtables of heart health stakeholders on a country-by-country basis in order to (i) identify context-specific roadblocks for patients; and (ii) promote use of a CVD scorecard tool to assess strengths and gaps in national CVD programs. All of this, he said, should lead to concrete national solutions and strengthened policies for improving cardiovascular disease. Additionally, the 2020 cohort of WHF’s Salim Yusuf Emerging Leaders Programme will be focused on diabetes, ostensibly resulting in three impactful research projects on health policy, systems, implementation, and science. Read on below for quotable quotes from the session’s panel. We note that some controversy is ongoing with IDF and we’re trying to determine what is happening with this organization – diaTribe’s partnership at present is specifically focused on WHF.

Quotable Quotes

• “If we ask our patient panel where they got their last diabetes prescription change, 70% are in primary care. And when primary care breaks down a visit, diabetes only gets half of the visit, and the other half is going to CV care because these patients are on statins and hypertension medications. So you have a 15-minute appointment, and there’s an opportunity there to glue it all together. We’re talking about bang for the buck: If you could get some of this into primary care and making earlier, better treatment decisions in primary care, that’s big.” – Richard Wood, dQ&A

• “In my role for the ACC, we talk about how disconnected the care is for diabetes. People are seeing 10 different specialists who are not talking, with different systems and different EHRs. How do we utilize tools we have today for comprehensive risk reduction? We need comprehensive lifestyle and medical therapies targeting aggressive goals for risk factors.” – Dr. Laurence Sperling, Emory Heart Disease Prevention Center

• “I think one of the most exciting components is how committed we are to including patient perspectives. That’s the key – probably most people with diabetes will never see this document though diaTribe will strive to spread the word! It’s important to ensure that the objectives, goals, strategies, and outcomes affect people with diabetes in their daily lives and experience living with diabetes. When we’re talking about roadblocks to implementation, let’s think about both patient and provider experiences. We really see that as a Venn diagram: there are shared roadblocks to implementation but also differences in how people with diabetes and providers experience the objectives we set out.” – Emily Fitts, The diaTribe Foundation

• “We have a lot of experience with cardiovascular disease prevention in diabetes in Europe, and we’ve found it’s a big issue – many patients with diabetes are not diagnosed. Among patients with coronary disease, only about 1/3 knew they have diabetes, another 1/3 had normal glucose tolerance, and nearly 40% of patients had diabetes and didn’t know. When we discuss the document and recommendation, we need better screening for diabetes.” – Dr. Kornelia Kotseva, National Heart & Lung Institute

• “Type 2 diabetes is a hidden disease – you know this. When we ask US patients if they have friends on Facebook with diabetes, none say they have friends with diabetes. This isn’t a visible disease. One thing that drives resistance to intensification, especially insulin, is that the disease becomes more visible. We have to identify and acknowledge those.” – Richard Wood, dQ&A

• “There’s a huge gust of wind in this area, and thinking about patient empowerment and awareness is going to ensure we’re not just a slight breeze in this gust of wind.” – Emily Fitts, The diaTribe Foundation

• “When you approach patients independently, they’re a lot more willing [to make diabetes changes] than physicians expect. 60% of patients are willing to do more, and we worked to filter bias out. There’s a huge gap, and we need to think about the disconnect between patient and physician perceptions.” – Richard Wood, dQ&A

• “In the ‘ACC Expert Consensus Decision Pathway on Novel Therapies for CV Risk Reduction in Patients with Type 2 Diabetes and Atherosclerotic Cardiovascular Disease,’ we have a section specifically dealing with how we can work better together. We refer to comprehensive diabetes care teams. When we talked about this in think tanks and roundtables, cardiologists were very comfortable with a heart team approach for cardiology patients. How do we connect that to care of people with diabetes? Our CV colleagues’ ears are perked to this intersection – it’s one of the hottest areas for growing knowledge, and they’re thirsty to implement this. We call for cardiology to be engaged and involved and champions of care.” – Dr. Laurence Sperling,

• “At the ACC Heart House Roundtable in December, we presented the results of a survey we did with dQ&A of about 1,000 people with diabetes, in which we evaluated awareness of the link between and cardiovascular disease and diabetes, as well as awareness of the cardio-protective benefits of some diabetes therapies. Although 72% of the respondents were already taking some sort of heart-related medicine, only 32% of those people said they “often think about their risk of heart disease.” That number dropped to 22% for people not currently on a heart-related medication. Even more surprisingly, about half of the respondents were taking a GLP-1 agonist and/or SGLT-2 inhibitor, but only 51% of that group felt strongly that there are diabetes drugs that can help protect you from heart disease, which again dropped to 28% among those not taking a GLP-1 agonist and/or SGLT-2 inhibitor. We also wanted to evaluate whether health care providers are discussing the risk of cardiovascular disease with their diabetes patients and if that is dependent on the type of provider. Less than a third of respondents had seen a cardiologist in the past year. Perhaps most shockingly, 37% of respondents had not seen a cardiologist not talked to a health care provider about their risk. As we know, people with diabetes are seeing PCPs for the most part, and even if they are seeing cardiologist, a lot aren’t talking about risk. It goes back to the question about systems – how can we create tools to empower both patients and providers?” – Emily Fitts, The diaTribe Foundation

• “You can’t push the end of a piece of string, so the challenge is how to identify patient perspectives, priorities, and preferences. If you think about asking someone to take an additional cardioprotective diabetes medication, people are already taking 10 pills a day in our panel. So you have to take one more, at a certain time of day, with or without food, or maybe an injection – that burden is significant. How are patients balancing that against benefits? It’s a complex decision, and not really even a conscious one sometimes…If you start to understand how people balance benefits and side effects, you can think about how to target meds to people with different priorities. It happens in the consumer research world all the time. We don’t treat patients like consumers but they are making consumer decisions. Part of the picture is having a way to assess patient priorities, and the key word is preferences.” – Richard Wood, dQ&A

• “One of our MD/PhD students is focusing on how, in all the consumerism out there, we don’t talk about dollars and cents. There is risk and benefit when it comes to medication, but really dollars and cents are decision makers, whether consciously or unconsciously.” – Dr. Laurence Sperling, Emory Heart Disease Prevention Center

6. ACC President Dr. Michael Valentine Says that Diabetes and Obesity Epidemics Will “Overwhelm the Capacities of the Developing World” in Opening Address

In his opening presidential address, Dr. Michael Valentine kicked off ACC 2019 by highlighting the impact diabetes and obesity have had on increasing the rate of cardiovascular disease and mortality worldwide. Pointing toward future challenges in the field, Dr. Valentine explained that rates of cardiovascular disease have been declining for several years but have recently seen an uptick. The cause? Dr. Valentine identified the growing diabetes and obesity epidemics as the driving force preventing the further decline of CVD prevalence. Notably, Dr. Valentine focused on the effect these related epidemics will have in the developing world. As an example, he cited staggering diabetes and obesity prevalence statistics in Brazil and Egypt (nearly two in five adults in Egypt are living with obesity – essentially equivalent to the US), and he ominously asserted that “these problems will overwhelm the capacities of the developing world.” To be sure, we’re glad to see the ACC prominently highlight the importance of addressing these epidemics in approaching cardiovascular disease management on a global scale. On this note, the ACC does host a thorough “Lifting the Burden of CV Disease and Diabetes” educational course on its website – we salute these efforts.

• Also of note, Dr. Valentine focused on issues of time allocation, isolation, and burnout in cardiology. He cited a concerning 2016 study in the Annals of Internal Medicine showing that cardiologists only spend 27% of their time face-to-face with patients. Nearly twice as much time is spent on EHR and desk work. Dr. Valentine channeled the frustration of many practicing cardiologists in his remarks, noting that “this is not what we spent six to eight years training for.” Such an impasse between the realities of practice and the ideal vision of clinical care is what drives feelings of isolation and burnout that are so prevalent in the field, according to Dr. Valentine. On this front, he struck an idealistic tone in urging the audience to remember the true reasons they are practicing in these seemingly inevitable moments of frustration.

7. Observational ATHENA Study Highlights Unmet Need Among Patients with Type 2 Diabetes and CAD, Real-World Applicability of THEMIS Results for AZ’s Brilinta

AZ presented data from the ATHENA study in two posters, focusing on disease burden and unmet need among patients with type 2 diabetes and CAD in the US using data from (i) the Diabetes Collaborative Registry (DCR) and (ii) the Optum Research database. These studies are meant to complement THEMIS, AZ’s CVOT for antiplatelet Brilinta (ticagrelor), for which positive topline results were released at the end of February. For context, THEMIS demonstrated the superiority of Brilinta + aspirin vs. aspirin alone for preventing MACE events in patients with known CAD and type 2 but without a history of MI or stroke (this type of therapy is already standard-of-care for post-ACS treatment), but no specifics on risk reduction or bleeding were presented in the topline release. Thus, these two observational, retrospective analyses speak to the real-world applicability and potential impact of THEMIS data.

• A THEMIS-like cohort (n=56,040) and a broader T2D-CAD cohort (n=69,790) were selected from the 1Q16 DCR dataset. Over a mean 1.2 years, MACE events occurred at a rate of 16.34 and 17.64 per 100 person-years, respectively (balanced across the three MACE components of MI, stroke, and CV death). Cumulative incidence of bleeding events was similar in both cohorts (0.13 events per 100 person years). For both cohorts, cumulative incidence of MACE events was higher in patients over 75 years old (~23 vs. ~13 for those <75 years), those with history of PCI or CABG (~17 vs. ~14 without), and those with multi-vessel disease (~18 vs. ~16). The researchers concluded that the high real-world event rates in both of these cohorts, despite widespread use of CV medications, suggests both that (i) THEMIS has broad real-world applicability to the type 2/CAD population; and (ii) patients with CAD and type 2 could benefit from improved CV risk reduction.

  • Both cohorts included patients ≥65 years old with type 2 diabetes, ≥1 dispensed fill of a glucose-lowering medication, and high risk of CV events defined as prior PCI, prior CABG, or ≥50% lumen stenosis of ≥1 coronary artery (or use of a P2Y12 inhibitor, ticagrelor’s MOA, in the T2D-CAD cohort). The THEMIS-like cohort excluded history of MI and any stroke, while the T2D-CAD cohort excluded history of MI, ischemic stroke, and intracranial bleeding.

  • At baseline, men comprised 63% and 61% of the THEMIS-like and T2D-CAD cohorts, respectively. Mean age was 74 years for both, proportion with history of PCI or CABG 73% and 58%, mean A1c 9.6% and 9.5% (76% missing data), and 91% and 92% were on an oral antiplatelet (including aspirin). 32% and 38% were receiving dual antiplatelet therapy with a P2Y12 inhibitor and aspirin.

• Results from the Optum Research database (both commercial and Part D) led to equivalent conclusions, though this analysis included patients ≥50 years old (other inclusion/exclusion criteria were similar). The THEMIS-like cohort (n=127,107) and T2D-CAD cohort (n=203,898) saw MACE event rates of 11.52 and 11.65 events per 100 person-years, respectively – slightly lower than those seen in the DCR analysis, but still quite high – over a median follow-up of ~2.9 years. Cumulative incidence of stroke was higher than of MI, which was higher than CV death incidence.

8. Comparative Real-World Outcome Study of >3 Million Type 2 Patients Finds No Increased Amputation Risk Associated with SGLT-2 Inhibitors

In a well-attended moderated poster session, Melbourne’s Dr. Sanjoy Paul presented findings from an EHR-based real-world study that found SGLT-2 inhibitors were not associated with increased risk of lower-limb amputations compared to other diabetes therapies. The study examined EHR data from the US and UK: Among just over 3 million patients, 99,891 were on an SGLT-2 inhibitor (no incretin background), 101,496 were on a GLP-1 agonist, 358,720 were on a DPP-4 inhibitor, and the remaining ~1.9 million were receiving other non-incretin diabetes therapies. Importantly, rates of lower-limb amputation in the SGLT-2 population were essentially identical to those in the GLP-1 and DPP-4 groups (HR=1.01, 95% CI: 0.66-1.54 and HR=1.31, 95% CI: 0.75-2.29, respectively). Conversely, the SGLT-2 inhibitor group actually showed a significantly lower rate of lower-limb amputations when compared to the non-incretin group (HR=0.62, 95% CI: 0.41-0.93), though confounding in this analysis seems more likely to us. We also note a very, very low event rate. Controversy over amputation risk associated with SGLT-2 inhibitors has raged on in recent years, following a significant signal driven by high-risk patients in the CANVAS program for canagliflozin. Importantly, no other long-term outcomes study for any other SGLT-2 inhibitor has demonstrated a similar significant increase in amputation risk. Still, real-world data on SGLT-2s and amputations remains mixed (see Dr. Cliff Bailey on the issue at Diabetes UK 2019): The real-world EASEL study found an increased risk of amputations with the SGLT-2 class (more patients were on Invokana rather than Jardiance or Farxiga), but OBSERVE-4D did not. At ESC 2018, a provocative observational cohort study did show a nearly doubled risk of lower limb amputations among patients starting on an SGLT-2 inhibitor vs. a GLP-1 – notably, we have heard some KOLs question the methodology of this study.

• No significant differences were seen in amputation rates across different SGLT-2 inhibitors (canagliflozin, empagliflozin, and dapagliflozin).

9. LEADER and SUSTAIN-6 Post-Hoc Shows Consistent Cardiorenal Benefits of Liraglutide and Semaglutide Irrespective of Diabetes Duration

University of Toronto’s Dr. Subodh Verma presented a new post hoc analysis of the LEADER and SUSTAIN-6 trials showing that both liraglutide and semaglutide (Novo Nordisk’s Victoza and Ozempic) deliver consistent benefits on cardiorenal outcomes across the spectrum of diabetes duration. Both liraglutide and semaglutide exhibited consistent benefits on MACE, expanded MACE, and nephropathy across patients with varying diabetes duration (p-interaction >0.05 for all endpoints; see figure below for full data). The Novo Nordisk-sponsored poster leverages this data to argue that these GLP-1s can be effectively used in a broad patient population of those with diabetes, from the newly diagnosed to those with long-standing type 2 diabetes.

In both trials, approximately half of all patients had a diabetes duration at enrollment of between 5 and 15 years, with the overall distribution similar between the two trials. For analysis, patients were binned by diabetes duration into <5 years, between 5 and 15 years, between 15 and 25 years, and greater than 25 years. As expected, increasing diabetes duration was associated with an increase in MACE event rate and other macrovascular complications.

-- by Ann Carracher, Martin Kurian, and Kelly Close