September 16-20, 2019; Barcelona, Spain; Day #1 Highlights – Draft

Hola from Barcelona! EASD 2019 is off to a fast start, with Monday solely dedicated to nearly 40 hours of industry symposia that our team has expertly weighed in on, as far as top highlights. And never worry, our full report will have lots more that we’re not including right this moment. Read on below, and see our preview for what’s in store the rest of the week!

Diabetes Therapy Highlights

1. Excitement Abound for DAPA-HF: KOLs Highlight “Transformative” Results, Note Disconnect Between CV Benefits and Glucose Lowering; Could Dapa-HF Help Overcome Inertia?

Across many of the industry symposia running during Day #1 of EASD 2019, we sensed tremendous excitement for the just-presented AstraZeneca Dapa-HF results, with multiple KOLs today commenting on the landmark nature of the trial. As a reminder, on the primary composite outcome of CV death/HF hospitalization/urgent HF visit, dapagliflozin 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) in HFrEF patients (both with and without diabetes, with consistent effects across both subgroups) in Dapa-HF. Today, Canada’s Dr. Subodh Verma characterized the trial as “transformative” and asserted that these results will “truly usher in a new day in the way we think of SGLT-2 inhibitors in the treatment of heart failure in the presence or absence of diabetes.” During multiple presentations over the course of the day, Dr. Verma came back to Dapa-HF time and time again, stressing the magnitude these results will have in shifting treatment paradigms in the space. We sensed that Dr. Verma wanted to convey the impressiveness of Dapa-HF results, especially in regard to the “high hurdle” it had to clear to show significant benefits: the benefit it demonstrated was on top of excellent standard of care therapy (no kidding, including 11% of patients that were on Novartis’ highly effective Entresto). Dr. Verma underscored that “it’s not always that you come across such a profound efficacy signal” – and we couldn’t agree more, especially considering the buzz that Dapa-HF generated at ESC 2019 where results were first presented. The fact that dapagliflozin still displayed such a robust effect beyond standard of care is what prompted Dr. Verma to posit that dapagliflozin will become a “foundational therapy” for heart failure, regardless of diabetes status. Whoa – over 25 million people globally are estimated to have heart failure and it’s mounting.

• Dr. Verma also believes that Dapa-HF will help solidify the fact that CV benefits associated with SGLT-2 inhibitors are completely decoupled from glucose lowering effects. A potential benefit of this is that it seems to be helping cardiologists overcome therapeutic inertia in their patients with ASCVD in whom they have been somewhat reluctant to prescribe SGLT-2s (and GLP-1s) to. Dr. Verma asserted: “Dapa-HF will help cardiologists overcome inertia in patients with ASCVD. We have patients with ASCVD who deserve these therapies, but the fact that these agents were presumed to be glucose lowering per se and agents that cardiologists should not dabble with, has been demystified after Dapa-HF. If anybody thinks that these effects are related to glucose after Dapa-HF, they need to think about that some more time.” We note that prescription rates of SGLT-2s and GLP-1s among cardiologists remain troublingly low (see Dr. Mikhail Kosiborod’s talk on this topic at ESC 2019 for more), and we are hopeful – actually, we’re optimistic - that Dr. Verma’s prediction on Dapa-HF turning the tide may come to fruition.

  • During a Lilly/BI symposium, Dr. Kausik Ray (a very highly regarded preventive cardiologist from Imperial College, London) emphasized a similar point. He noted that Dapa-HF demonstrates the disconnect of its mechanism of action in terms of CV benefit from glucose lowering: “Glucose lowering is important but for this endpoint it’s not really relevant. What we should be thinking about is organ preservation and improving outcomes.” From our view, of course glucose-lowing is important, but it’s not the mission of cardiologists – allowing that to be a bonus is more than fine.

• University of Glasgow’s Dr. Mark Petrie echoed many of Dr. Verma’s sentiments while also pointing out that these results should prompt greater collaboration between cardiologists and diabetologists. Dr. Petrie noted that since not all the patients in Dapa-HF had diabetes, there must be more involved teamwork between diabetologists and cardiologists about thinking how to fit these traditional “diabetes drugs” into cardiology. He struck a hopeful tone, noting that greater collaboration could yield a better understanding of how these impressive benefits are conveyed, along with ensuring that the right patients are benefitting from these treatments. On a separate note, Dr. Petrie also underscored the reassuring safety data from Dapa-HF: “When I talk to diabetologists, there’s some concern about adding dapagliflozin to ‘sick’ patients that are already taking a lot of other medications. The patients in Dapa-HF were these ‘sick’ patients. Despite this, there were no adverse safety data – this is very, very reassuring data in the use of this drug in ‘very sick’ patients.”

Fortifying this enthusiasm for Dapa-HF was the news Monday morning (the timing could not have been better) that FDA has granted a Fast Track designation to Farxiga in heart failure. Our understanding is that this designation allows for more frequent communication between FDA and AZ, as well as for rolling review of any future sNDA submission (meaning that FDA would review application materials as they are ready, rather than waiting for the complete application). A Fast Track designation also means that if Farxiga meets certain eligibility criteria in clinical findings, it could be granted Priority Review for an expedited approval process (we think this might be likely given the landmark nature of Dapa-HF results). This is a huge win for AZ and also comes on the heels of Farxiga earning a similar Fast Track designation in CKD earlier this year. Elsewhere in the SGLT-2 inhibitor in heart failure landscape, Lilly/BI’s Jardiance gained a Fast Track designation in chronic heart failure in July 2019 – ultimately, this new designation is a positive for the class.

SGLT-2 Inhibitors in Heart Failure Landscape: Ongoing Outcomes Trials

2. SGLT-2s or GLP-1s for Renal Protection? KOLs Compare Data Between Classes; Currently Favor SGLT-2s, But Future Data on GLP-1s (Namely Semaglutide’s FLOW Trial) Could Shift Debate

In reviewing renal outcomes data for SGLT-2 inhibitors and GLP-1 agonists, Dr. David Cherney (Toronto General Hospital Research Institute) expressed more confidence with SGLT-2s given their impressive benefits in outcomes trials on hard renal endpoints. Dr. Cherney asserted that renal outcomes data for the GLP-1 class is “very interesting,” as the class has been shown to reduce renal composite endpoints that include reducing macroalbuminuria. However, the effect of the class does appear to be driven by this endpoint, as removing macroalbuminuria considerations for these composite endpoints results in no significant benefits with GLP-1s. Dr. Cherney noted that it’s “still an open question that needs to be addressed in dedicated trials” and pointed to Novo Nordisk’s FLOW trial investigating semaglutide on renal outcomes as an important step for the field in helping to answer this question. As a reminder, the FLOW trial was initiated in June 2019 and will enroll 3,160 patients with type 2 diabetes and CKD (eGFR 25-75 ml/min/1.73 m²) for up to five years. Expected completion is August 2024, and the primary endpoint is a composite of persistent eGFR decline of ≥50%, reaching ESRD, and CV or renal death. Traditional MACE endpoints will also be measured and we hope Time in Range and PROs will be also. This is a highly notable, first-ever renal outcomes trial among GLP-1s, and it’s clear that enthusiasm is high for the trial’s potential in further informing renal benefits of the GLP-1 class. We’ve sensed optimism from Novo Nordisk that more efficacious GLP-1s (such as semaglutide) may be able to compete in the future with renal outcomes data from the SGLT-2 class, along with providing the additional benefit of not losing their glucose lowering abilities in patients with a lower eGFR.

• Dr. Cherney’s assessment of the GLP-1 class on renal outcomes aligns with broader KOL consensus that we’ve gathered on the topic. In light of REWIND results, KOLs appeared a bit underwhelmed by dulaglutide’s efficacy on renal endpoints, and noted that data for the GLP-1 class on these endpoints somewhat pale in comparison to those shown by the SGLT-2 class, to date.

• Later on in the day, University of Helsinki’s Dr. Per-Henrik Groop made this same point in a Lilly/BI sponsored session, noting that GLP-1 renal outcomes data is primarily driven by macroalbuminuria reductions and not on the same hard renal outcomes that the SGLT-2 inhibitor class has so impressively shown benefits on.

• Dr. Hiddo Heerspink also weighed in on the subject during AZ’s symposium, asserting that if he had to choose between a GLP-1 and an SGLT-2 in terms of protecting kidney health in a post-ACS type 2 diabetes patient, he would surely choose the SGLT-2.

• Dr. Cherney also commented on the usage of SGLT-2s in patients with low eGFRs, asserting that renal-protective effects are likely preserved across all eGFR levels. Indeed, although SGLT-2 inhibitors remain contraindicated for those with severe renal impairment (eGFRs below 45 mL/min/1.73 m²), their benefits on renal outcomes have been significant in those with lower eGFR levels. Dr. Cherney noted that CANVAS, EMPA-REG, DECLARE, and CREDENCE all enrolled patients down to an eGFR of 30 (with benefit still seen here), while ongoing dedicated CKD trials in Dapa-CKD and EMPA-KIDNEY are going down even further in terms of eGFR (25 and 20, respectively). Of course, the current contraindication for SGLT-2 inhibitors in those with eGFRs below 45 is not because of potential safety concerns, but because of the attenuation of glycemic benefits at lower eGFRs. Dr. Cherney predicted that the future of SGLT-2 inhibitors will include broader use of the class in these patients with lower eGFRs, seeing as their impressive effects on renal and CV outcomes are significant in these patients. We’re curious whether results from Dapa-CKD and EMPA-KIDNEY (along with those from CREDENCE) may help regulatory agencies reconsider this contraindication. Dr. Cherney appeared confident on this front, noting that it’s likely that these labels will be changed to recommend usage in patients with eGFR’s as low as 30 (he also commented that this is already the case in Canada).

3. Dr. Subodh Verma: Patients at Risk for Heart Failure Should Be Treated “Before the Horse Leaves the Barn” – No Need to Confirm Heart Failure Diagnosis Before Initiating Treatment

Jumping into a conversation regarding when targeted evaluations for heart failure might be initiated in an at-risk patient, Dr. Subodh Verma stressed that it’s not necessary to actually confirm a heart failure diagnosis before using SGLT-2 treatment. After all, the trials showing robust benefits in heart failure prevention (namely DECLARE) did not discriminate treatment for those with heart failure at baseline; instead, the trials were designed to show that at-risk patients could be treated and heart failure hospitalizations could be prevented without an actual heart failure diagnosis. Said Dr. Verma: “These patients are at risk for heart failure and they should be treated before the horse leaves the barn. That’s how the trials were done and that’s how the treatments should be used. Scientifically speaking, saying ‘let me find some heart failure before I start treating’ is a disservice to our patients.” This response came in regard to an ongoing discussion of how to better identify heart failure in clinical practice, and what steps may be necessary to make more targeted evaluations (such as completing an ECG). Dr. Verma felt that these steps may be unnecessary in the context of deciding whether to use an SGLT-2 or not and made an analogy to the use of statins not requiring such diagnosis before their use. We’re glad to see Dr. Verma emphasize this point, since we imagine that with data emerging on both the prevention (through DECLARE) and treatment (through Dapa-HF) of heart failure, there may be some ongoing confusion on how to best initiate and use these therapies and in which patients.

4. Dr. Richard Bergenstal on Beyond A1c: TIR Linked to Risk of Complications, AGP is the “EKG for Glucose Management”

In a typically impassioned keynote during Novo Nordisk’s symposium, Dr. Richard Bergenstal (Park Nicollet) urged for glucose management beyond A1c, specifically looking to CGM results as the best alternatives. Notably, many points from the presentation echoed themes from his past talks on CGM at AACE 2019, IDC 2019, and EASD 2018. We were particularly struck by how Dr. Bergenstal discouraged the paradigm in which A1c goals are lowered in hopes of reducing hypoglycemia; instead Dr. Bergenstal asked, “Can’t we do both? Can’t we have a good A1c without hypoglycemia?” – we surely agree that it’s time to raise the bar and do both! This call-to-action felt especially relevant considering Novo Nordisk’s recent partnership with Medtronic, rounding out the company’s portfolio of collaborations with all major CGM makers. 

• Looking to the future, Dr. Bergenstal decreed, “CGM will never win the day if you are not convinced that it is related to long term complications.” The implementation of A1c as a “gold standard” primarily comes from its successful correlation with the relative risk for co-morbidities such as retinopathy, nephropathy, and neuropathy. Dr. Bergenstal explained that TIR must be shown to do the same for it to be validated as the next “gold standard.” To this end, Dr. Bergenstal et al. published a 2018 paper in Diabetes Care that linked time-in-range with complications, including retinopathy and microalbuminuria, using DCCT data. Notably, that paper used a quarterly measured 7-point fingerstick profile to estimate time-in-range, rather than a CGM. With a CGM, this correlation between TIR and complications could be even stronger.

• We especially enjoyed the parallel Dr. Bergenstal drew between EKG outputs and AGPs: “the EKG for glucose management,” highlighting the comprehensive and concise nature of both reports. While there is a plethora of EKG machines available, they all generate a consistent output that includes (i) metrics that answer “is there a problem?”; and (ii) a profile that answers “where is the problem?” AGPs were created to do the same, and Dr. Bergenstal guaranteed that “within one minute of explaining, a patient will be able to understand how to read an AGP.” To finish, he pointed to the ADA’s August update to its 2019 Standards of Care that added standardized TIR goals and CGM metrics, as well as recommended use of the one page AGP.

• Dr. Bergenstal explained that A1c has been the “gold standard” of diabetes management for the last 26 years, but that the measurement carries a number of limitations, including (i) it does not identify hypoglycemia or glucose variability; (ii) A1c has accuracy problems for ~15% of the population (who have hemoglobinopathies, iron deficiency, kidney disease, etc.); and (iii) there is no broad agreement on a target value (AACE recommends <6.5%, ADA <7%, and ACP <8%). To the first point, Dr. Bergenstal explained that while many of his patients have the same A1c values, their % time in hypoglycemia (<70 mg/dl) can vary almost 10-fold! Perhaps most importantly, Dr. Bergenstal underscored that the current usage of A1c alone to guide treatment is not working, citing rising global A1c values, despite it being used as the foremost regulatory measure.

• Lastly, Dr. Bergenstal shared two of the most helpful acronyms he uses to guide his patients understanding of their AGP: (i) MGLR: more green, less red; and (ii) FNIR: flat, narrow, and in range. “More green” being more time in target range, and “less red” being less time below target range, as shown at the top of an AGP. We’d love to see specifically less “yellow and red” since hyperglycemia is connoted by yellow.

5. Merck Symposium Highlights SGLT-2 Inhibitor Steglatro; Anticipation for Early 2020 VERTIS CV Readout

A Merck-sponsored symposium provided a deep dive on the company’s portfolio of oral diabetes therapies: Pfizer-partnered SGLT-2 inhibitor Steglatro (ertugliflozin) and DPP-4 inhibitor Januvia (sitagliptin). Dr. Sam Dagogo-Jack (University of Tennessee Health Science Center, Memphis, Tennessee) provided a deep dive on the ongoing and highly-anticipated CVOT for Steglatro, VERTIS CV. The trial’s design has several notable features: (i) The trial is strongly positioned as a secondary prevention study, enrolling 99.9% of participants with established CVD; (ii) as we learned last year at EASD 2018, 23% of the VERTIS CV patient population has a history of heart failure, bolstering the statistical power to investigate this crucial secondary outcome. To this end, we believe VERTIS CV is also the only SGLT-2 CVOT to collect baseline heart failure status in all participants, further boosting the study’s ability to examine the increasingly apparent link between SGLT inhibition and heart failure outcomes. Dr. Dagogo-Jack confirmed that enrollment is complete (>8,000 participants) and the study is projected to complete in December 2019, with results available as early as January 2020 (at one point it was as early as 4Q19). Commenting on the wave of positive SGLT-2 CVOTs in general, Dr. Ofri Mosenzon (Hebrew University Hospital, Jerusalem, Israel) provocatively argued that we should start referring to these agents as “disease-modifying therapies” for their ameliorating effect on the natural history of crucial diabetes complications like heart failure and CKD. Dr. Daniel Drucker (University of Toronto, Canada) contextualized that this is “something that five years ago none of us could have imagined” before the paradigm-shifting EMPA-REG OUTCOME readout at EASD 2015. Whew!

• Turning to Merck’s DPP-4 inhibitor franchise, Dr. Juan Frias (National Research Institute, Los Angeles, CA) highlighted the CompoSIT trial program. CompoSIT-R found superior A1c-lowering with Januvia (sitagliptin) over AZ’s SGLT-2 inhibitor Farxiga (dapagliflozin) in patients with mild renal impairment (to the tune of -0.15%). This makes sense given that the glucose-lowering mechanism of SGLT-2 inhibitors is strongly dependent on renal function. CompoSIT-I found greater reductions in A1c and FPG when sitagliptin was continued vs. withdrawn at the initiation of insulin glargine therapy, with no increase in hypoglycemia. Finally, CompoSIT-M showed A1c benefits for adding sitagliptin during metformin uptitration vs. metformin alone. Together, these trials show the relevance of Januvia for three common clinical situations: (i) As an alternative to SGLT-2 inhibitors in people with renal impairment, (ii) as complement to insulin therapy (that boosts the glucose lowering with no added risk of hypoglycemia), and (iii) as an addition to intensifying metformin therapy (sooner rather than later, to combat clinical inertia).

• Dr. Drucker kicked off the symposium with some words of wisdom for clinicians eyeing exciting advances in beta cell basic science. “Let’s apply the same stringent criteria to basic science that we do in the clinical arena,” he advised. Highlighting the poor replicability of early basic science studies (given methodological differences) and imperfect translation from mice and cell culture to humans, Dr. Drucker implored the audience not to “put too much faith in the latest, greatest basic science finding.” With exciting clinical trials, Dr. Drucker pointed out that people are quick to acknowledge that the effect may not hold up in subsequent, larger and differently-designed trials, or for other members of the same drug class. He argued that the same amount of nuance should be applied when interpreting basic science findings.

6. Novo Nordisk Symposium Highlights GLP-1 + SGLT-2 Combo Therapy and Lack of GLP-1 Penetration – Could Oral Sema Move the Needle?

GLP-1s stole the show at Novo Nordisk’s packed corporate symposium on moving beyond the standard in diabetes and obesity care, led by a star-studded cast including Drs. Rory McCrimmon, Subodh Verma, Ofri Mosenzon, Thomas Wadden, Melanie Davies, Vanita Aroda, and more. Themes from the session included (i) optimized use of GLP-1s for patients with type 2 diabetes; (ii) optimized use of GLP-1s for patients with obesity; and (iii) the future of GLP-1s. Read on for some top highlights below!

• Dr. Rory McCrimmon focused on the potential benefits of combining GLP-1s and SGLT-2 inhibitors for patients with type 2 diabetes, displaying a particularly helpful table of the two drugs’ possible counterbalancing effects. For example, hepatic glucose output is negatively impacted by the use of an SGLT-2 inhibitor; however, adding a GLP-1, which beneficially decreases glucose production, can neutralize the effect for patients. Dr. McCrimmon cited the AWARD 10, DURATION 8, and LIRA-ADD2SGLT-2i clinical trials as evidence of the combination’s success.
  

• Dr. Thomas Wadden and Dr. Ofri Mosenzon tackled GLP-1 usage in patients with obesity – the former focusing on intensive behavior therapy (IBT) and the latter on combination usage with basal insulin. Dr. Wadden presented data from the SCALE IBT trial (n=282), which combined liraglutide 3.0 mg with IBT (prescribed exercise of 250 min/week, reduced caloric intake based on body weight, and 23 IBT counseling visits) for patients with a BMI ≥30 kg/m² but without diabetes. In the study, behavioral counseling with placebo induced a mean reduction in baseline weight of 5.4% at six months, confirming the effectiveness of brief IBT alone. Although six-month results did not differ greatly for those additionally on liraglutide, patients who stayed on the therapy at week 56 saw an impressive loss of 9.1% of baseline weight. Dr. Mosenzon focused on the SCALE Insulin trial (n=396) which looked at liraglutide 3.0 mg + IBT vs. placebo + IBT in patients with overweight/obesity and basal insulin-treated type 2 diabetes. Cumulative results showed that patients also on liraglutide were 3.3x more likely to achieve ≥5% weight loss, A1c <7%, and no documented symptomatic hypoglycemia (p=0.0006) – these results are incredibly encouraging, and overall, we see tremendous potential for GLP-1s in the evolving field of diabetes treatment.

• In terms of the future of GLP-1s, oral semaglutide was the obvious center of attention. While the overall audience opinion on oral semaglutide was generally positive (praise for the PIONEER trials received light but spontaneous applause), the question of pricing, and more broadly why only ~10% of eligible patients are on GLP-1s, did seem to be a hot topic from the audience. In response to this query, both Dr. Melanie Davies and Dr. Vanita Aroda pointed to lack of provider comfortability as a major factor. Dr. Davies noted that while DPP-4 inhibitors are similarly priced to SGLT-2 inhibitors in her native UK, DPP-4 inhibitors are still far more often prescribed, demonstrating that price is not the only issue. Dr. Aroda added that some practitioners feel less prepared to put their patients through careful dose escalation and explain GI side effects compared to other drugs. We’re hopeful that oral semaglutide could help to bend the curve here by reducing injection burden (perceived or real). To end Q&A, the panel collectively agreed that the challenges with GLP-1 accessibility are not just in the United States and Europe but also in Africa and Asia – we wholeheartedly agree and hope that the conversation on access is not just limited to those who have the means to speak about it!

7. Across-T2D Website Launched: Educational Resource for HCPs Outside US and UK; Slides From Major Trials Impacting Clinical Care Posted to Bridge Information Gaps

Dr. Bernie Zinman highlighted the launch of www.Across-T2D.com – an informational website that will provide educational resources for international HCPs (outside the US and UK) that care for people with diabetes. Notably, Dr. Zinman explained that slides from most major clinical trials and other important emerging data that impact clinical practice will be published on the site for HCPs to access. (Even without restricted access, we were able to see slide decks for the presentation of DECLARE-TIMI 58, HARMONY Outcomes, and CARMELINA—cool!). The project is supported by Lilly/BI and a whole host slew of associated HCPs on the site’s steering committee. The website is intended to be a living resource that is continuously modified and updated with the latest information so that HCPs can stay up to date with the latest data that impacts clinical care. We’re thrilled by this move that should help bridge knowledge gaps between the latest clinical advances and HCPs, and wonder when this tool might be made available to HCPs in the US and UK as well.

8. Novartis Symposium Highlights Heart Failure; Promise for SGLT-2s and Entresto in HFrEF; Unmet Need Persists in HFpEF

A Novartis-sponsored symposium shed light on the increasingly apparent intersection of diabetes and heart failure. Expectedly, a major focus of discussion was Novartis’ heart failure therapy, Entresto (sacubitril/valsartan), indicated for HFrEF. Great interest has surrounded the question of whether Entresto’s benefits extend to the increasingly common HFpEF, which is thought to be more tightly linked to adiposity, obesity, and diabetes. Sadly for the diabetes and cardiology community, hot-off-the-presses data from the newer PARAGON-HF trial indicate that this may not be the case. Entresto narrowly missed significance on the primary endpoint of reducing CV death and total HF hospitalizations (p=0.06), and subgroup analyses for hospitalization for HF and HF death were also not significant. Positive results in PARAGON-HF would have represented the first ever for a potential treatment of HFpEF, and the two cardiologists on the panel, Dr. Martin Cowie (Imperial College, London, UK) and Dr. Andriaan Voors (University Medical Center Groningen, Netherlands), took this as an opportunity to underscore the unmet need in this form of heart failure.

• Following promising signals for heart failure prevention in recent CVOTS, SGLT-2 inhibitors are also a source of great hope for the treatment of heart failure, particularly in people with diabetes. To this end, Leicester’s very highly regarded Dr. Melanie Davies highlighted the DAPA-HF trial, which stole the show at this month’s ESC meeting with a 26% risk reduction on CV death and hospitalization for heart failure with AZ’s Farxiga (dapagliflozin) in HFrEF patients with and without diabetes. These results solidified dapagliflozin as the first SGLT-2 inhibitor to show benefit in heart failure, and the first SGLT-2 inhibitor to demonstrate beneficial effects in a large-scale trial of people without diabetes. On this wave of excitement surrounding SGLT-2s, we note that two additional trials are ongoing to address HFpEF specifically: EMPEROR-Preserved for Lilly/BI’s Jardiance, DELIVER for AZ’s Farxiga.

• On the epidemiological front, Dr. Clifford Bailey (Aston University, Birmingham, UK) provided an important reminder of why these trial results matter so much: The combination of heart failure and diabetes is a “deadly duo.” Diabetes increases the risk of heart failure, and heart failure in turn increases the risk of diabetes. The combination of the two is associated with high mortality: According to data from the Swedish Heart Registry, median survival for patients with HF and diabetes is only 3.5 years (vs. 4.6 years for HF alone). Making matters worse, therapy for heart failure and diabetes can sometimes counteract each other. Dr. Bailey pointed out that thiazides, statins, and beta-blockers can potentially increase blood glucose, whereas diabetes therapies (certainly TZDs and potentially SUs and insulins) can negatively impact heart failure prognosis. This puts the importance of SGLT-2 inhibitors into perspective, with their beneficial impact on both diabetes and HF.

Diabetes Technology Highlights

1. Dexcom Decision Support Pipeline: InPen Data Now Integrated with Clarity; In- Development Bolus Calculator (Launch Before G7?), Hypo Prediction, and More

Dexcom CTO Mr. Jake Leach closed out Dexcom’s afternoon symposium with a look at the company’s decision support pipeline and new data on the positive correlation between Dexcom remote monitoring usage and time-in-range in adults. On the same day as smart pen partnerships were announced between Abbott and Sanofi and Medtronic and Novo Nordisk, we learned that the smart pen/CGM data integration between Companion Medical’s InPen and Dexcom’s Clarity has successfully launched after its announcement in June. The combined view of insulin data and CGM data (see below) is going to be valuable for facilitating smarter conversations and adding more context over CGM alone. Dexcom also continues to develop its decision support software for injection users, including a “CGM-informed” bolus calculator, enhanced hypoglycemia prediction, “on-demand” sleep and exercise advice, and ongoing glucose control assessment. In Q&A, Mr. Leach revealed that Dexcom will try to get the bolus calculator out before the launch of G7 (currently targeted for late 2020-early 2021). Otherwise, there were no new updates or screenshots relative to those shared at the December 2018 Investor Day. In the smart pen arena, Dexcom also has partnerships with Novo Nordisk (expected 2020 launch of its smart pens) and Lilly (submitted to the FDA as of 1Q19).

• Dexcom presented new real-world data showing a slight correlation between remote monitoring and time-in-range in US adults. Over half (53%) of the 50,000 adult G6 users analyzed used the remote monitoring feature, with higher utilization in the younger age group (18-24 years). We think this should be closer to 100% and look forward to helping move it that direction – we don’t know how many use Clarity (full stop) but that should indeed be 100%. Users with more followers had slightly lower mean estimated glucose values and slightly increased time-in-range, which we found interesting. The absolute differences were quite small, with a roughly 5%-point difference in time-in-range and a <10 mg/dl difference in mean glucose (see the charts below). Number of followers was also correlated with lower percentage of time spent in hypoglycemia (<70 mg/dl). We note that these are correlations, and thus subject to an important confounding bias – number of followers may simply be an indicator of those with more support or those who are more engaged, rather than a causal driver of outcomes. Earlier this year, the same analysis was performed on G6’s pediatric users (2-18 years) and found stronger correlations between number of followers and improved glycemia.

2. CME on CGM: Favorable Shift in Audience Attitudes Toward CGM, Esp in T2D; Reviews of Registry, RCT, and Real-World CGM Outcomes Data

During the course of an Abbott-sponsored CME session on CGM, audience attitudes shifted quite dramatically in favor of the technology. Before the talks, about half of the audience (51%) said that they currently recommend/prescribe CGM for >60% of people with type 1 diabetes; following the talks, that rose to nearly two-thirds of the audience (64%) – a pretty quick gain after seeing some of the evidence. The lift was even bigger for type 2 diabetes: before the talks, 17% of the audience said that they recommend/prescribe CGM for >60% of type 2s on insulin; this rose to 41% by the end. (And in line, 28% said that they recommend/prescribe CGM for <5% of this population, which dropped to 3% after the talks, which we were very glad to see.) Of course, there was some turnover in the room and therefore those participating in the poll, but at face value, this marked shift highlights the degree to which knowledge about CGM and its evidence-based benefits is lacking, even among conference-going diabetes care providers. As a field, we have to continue beating the drum and spreading awareness – particularly among primary care providers – so that we can further increase uptake globally. University Hospitals of Derby and Burton’s Dr. Emma Wilmot put a finer point here: “I worry we’re in a situation where history is repeating itself [referencing the 20-year lag between the recognition of SMBG’s value and it replacing urine glucose strips]. We have a range of sensor-based CGMs available, and it’s clear that insulin dosing is easier, hypos are less frequent, and a vast number of patients prefer it. I don’t want to wait two decades before they’re replaced.”

• What brought about the shift in attitudes among the audience? A well-put-together walkthrough of registry data from Hospital General de Segovia’s Dr. Fernando Gomez-Peralta, a review of landmark trials and outcomes beyond A1c from Hannover’s Prof. Thomas Danne, and real-world evidence from Dr. Wilmot. Dr. Gomez-Peralta detailed analyses of the latest batch of T1D Exchange data showing that A1c is lower for CGM users across all ages (Rodbard, DTT 2019), independent of insulin delivery modality (Foster et al., DTT 2019). To these data, he added a recent BDC analysis of early CGM initiation in newly diagnosed type 1s (lower A1cs, fewer diabetes-related ED visits), as well as registry data from Spain showing that A1cs are similarly lower in those on CGM vs. SMBG across all ages, with the exception of ages 14-17. Prof. Danne reviewed the pediatric BEAGLE study of FreeStyle Libre in the UK, DIaMOND and GOLD (Dexcom G4 in MDI), IMPACT (FreeStyle Libre in type 1), REPLACE (FreeStyle Libre in type 2s on basal-bolus insulin), and CONCEPTT (Medtronic CGM in pregnant women), all of which demonstrated strong benefits of CGM. He also contributed a couple of exuberant beyond-A1c quotes (bullets below). Finally, Dr. Wilmot pointed to real-world CGM outcomes, starting with the ABCD UK FreeStyle Libre audit (presented at ADA 2019), showing that individuals who started on FreeStyle Libre decreased A1c (median -0.6% from baseline 8.2%) with impressive concurrent drops in DKA and hypoglycemia-related hospital admissions. Dr. Wilmot stressed that these results are encouraging out to six months, but must conclusion must wait until 12-month follow-up data are available. She also highlighted a week-old publication from an Edinburgh prospective observational study, in which A1c was tracked in 900 type 1 FreeStyle Libre users. Overall median A1c change at a median follow-up of ~eight months was -0.4% (baseline: 7.9%). The proportion able to achieve A1c <7.5% at follow-up increased from 33% to 51% in the FreeStyle Libre arm, and the proportion with A1cs >9% dropped from 18% to 9%. Unsurprisingly, low baseline frequency of SMBG and higher baseline A1c were predictive of an A1c drop >0.5%. However, previous participation in structured diabetes education (DAFNE) was not predictive of success, as measured by A1c. Dr. Wilmot posited that this is because “CGM in itself can be an educational tool – people report knowing more about diabetes than they ever have before just by using it.”

  • “We are really at a tipping point. Now that these systems are getting better and better, how long are we going to buy blood glucose strips? I might also ask how many of you are buying CDs still vs. how many of you are streaming music? I’m trying to tell you how quickly a change is happening…My prediction is we won’t be using BGMs for very much longer.” – Prof. Danne

  • “Very often I have the feeling that the biggest problem with sensor accuracy is in the heads of our patients. They are used to point accuracy….they don’t realize that we have to move away from point accuracy toward trend accuracy, and we have to judge the glucose according to a trend. With a trend, we have new information. Of course, MARD is important, but in my book, there really isn’t that much of a difference anymore.” – Prof. Danne

3. Dr. Polonsky: Diabetes Devices Must Help PWDs Feel In Control, Not Control Them; People Are Often Ambivalent About Receiving Diabetes Data, Advice, & Support

After acknowledging cost and access, Behavioral Diabetes Institute’s Dr. Bill Polonsky framed CGM uptake and continued use as hinging on two “ambivalences” that everyone with diabetes feels:

• Ambivalence about receiving data. “On the one hand, alerts, alarms, and as much data as possible are great and help me stay on track. But at the same time, I’m sick and tired of thinking about diabetes and I want to be bothered by alarms as little as possible.”

• Ambivalence about receiving advice and support. “On the one hand, leave me alone; I don’t need anyone to bother me about diabetes. But at the same time, don’t leave me alone with diabetes, I could really use your support and encouragement.”

Therefore, he said, most want a device that (i) alerts them as needed with critical alarms and never bothers them with alarms (inaccurate/frequent alarms are an oft-cited reason for discontinuation); and (ii) provides guidance and encouragement, but also leaves them alone. To resolve this, people with diabetes will find a point of compromise between how well a device supports and helps them feel in control vs. how much it controls them. For example, those who are insulin-dependent are more likely to feel in control when they can feel safe from hypoglycemia. For people who are at less risk of hypoglycemia, feeling in control may be having the ability to choose when/if they will be informed or bothered. He gave a nod to the FreeStyle Libre 2 with optional and customizable threshold alarms as a step forward to meeting both “ambivalences” and predicted that the future will hold systems that allow for “even more profound, personalized control over alert and alarm settings” as well as more opportunities to help people understand how to use alarms optimally for their personal preferences and attitudes. He concluded that it’s “clearly most important” for healthcare providers to discuss with patients the pros and cons of different devices as they weigh their options. One way he believes providers can help patients make the best technology decisions is by identifying subgroups, such as those described by Tanenbaum et al. Underlying the whole talk was the idea that CGM is the future for all persons with diabetes in some form, but it is only effective when used, and most effective when used most frequently!

4. Dr. Peter Jacobs: Dexcom G6 “Could be Used Safely” with Aerobic, Resistance, and High Intensity Interval Exercises

OHSU’s Dr. Peter Jacobs presented unpublished data on accuracy of the Dexcom G6 during exercise, finding that the mean absolute relative difference (MARD) between G6 and a Contour Next BGM reference did not significantly change through six time-points during and thirty minutes after exercise. The study was performed on 24 adults with type 1 diabetes, ages 18-50, who performed two in-clinic study sessions of aerobic, resistance, and high intensity exercises. Across all time points and exercises, MARD never rose above 15.9%, relatively impressive given the difficulties in measuring glucose values during exercise (rapidly changing glucose, fluid distribution, and lag between blood and interstitial glucose levels). On a Clarke Error Grid, 99.3% of measurements fell within the A and B regions. The G6 had a negative mean relative difference (MRD) across all types of exercise at most time-points, meaning the CGM values tended to be lower than the BGM readings. This negative bias is certainly better than a positive bias (CGM values higher than actual values), which could lead to hypoglycemia.

Mean absolute relative difference (MARD) and mean relative difference (MRD) between CGM and BGM readings during and after three types of exercise. Aerobic and resistance exercise lasted 30 minutes and high intensity interval (HII) exercise lasted 25 minutes. Yellow highlights denote statistical significance.

• Dr. Jacobs also found that glucose values dropped significantly over all three exercises, with a mean 48 mg/dl drop during aerobic exercise, 36 mg/dl drop during resistance exercise, and 11 mg/dl drop during interval exercise based on BGM readings. Falling blood glucose during exercise is a certainly known phenomenon, but Dr. Jacobs emphasized the heterogeneity of responses during exercise by showing the BGM and CGM traces for all 24 study participants. In all three types of exercise, participants had vastly different glucose changes.

• The Dexcom G6 outperforms FreeStyle Libre and previous Dexcom CGMs on accuracy during exercise. A paper from earlier this year in Diabetes Medicine found a median ARD of 22% in ten people with type 1 diabetes performing aerobic exercise. A study of Dexcom G4 with 22 people performing aerobic exercise found median ARD of 16.8% during exercise with an average +14 mg/dl bias during the exercise session. These comparisons do feel a bit artificial at this stage, since no one is taking frequent fingersticks during exercise – for those who want to stay in-range during exercise, CGM is a must!

--by Albert Cai, Ursula Biba, Rhea Teng, Abigail Dove, Brian Levine, Martin Kurian, and Kelly Close