American Diabetes Association 78th Scientific Sessions

June 22-26, 2018; Orlando, FL; Day #5 Highlights – Draft

Executive Highlights

  • This year, ADA really saved some of the best for last. Tuesday morning started with a presentation of ADA/EASD’s new draft consensus statement on type 2 diabetes treatment selection. This draft is open for public comment until July 2, 11:59 pm EDT, and we highly recommend watching the full presentation once it goes live here before sending comments to Major recommendations include GLP-1s for patients with ASCVD, SGLT-2s for those with heart failure, and, notably, considering both GLP-1 + SGLT-2 together if one alone isn’t enough. Importantly, HCPs are to assess an individual’s CV status as step #1 in diabetes care. Liraglutide is preferred over semaglutide is preferred over exenatide once-weekly, while empagliflozin is preferred over canagliflozin. Metformin is still recommended as first-line therapy, although, also notably, Dr. John Buse (who co-chaired this writing committee for ADA/EASD alongside Prof. Melanie Davies) called this into question during his concluding remarks. Moreover, the writing committee has not found sufficiently robust evidence for initial combination therapy (we assumed it would at a specific A1c threshold), nor has it suggested SGLT-2s (or GLP-1s) for their renal protective properties (and during Q&A, Dr. Silvio Inzucchi called this into question).

  • Tuesday offered even more content on SGLT inhibitors in type 1 diabetes: Dr. Julio Rosenstock presented topline data from Lilly/BI’s EASE-2 and EASE-3, investigating Jardiance (empagliflozin) as adjunct to insulin in 1,680 type 1s. EASE results seem largely in line with findings from DEPICT (AZ’s dapagliflozin) and inTandem (Sanofi/Lexicon’s sotagliflozin), meaning the treatment offered additional A1c-lowering, weight loss, and time-in-range. Very notably, Dr. Rosenstock pointed to a 2.5 mg dose of empagliflozin that conferred no DKA risk; this is a much lower dose than Jardiance (10 mg or 25 mg), but apparently, and unexpectedly, efficacy with 2.5 mg was on par with that from 10 mg. Is this too good to be true? We’re waiting eagerly for the full data, expected at EASD. Later in the same symposium, Dr. David Cherney outlined reasons we need a CVOT of SGLT inhibitors in type 1 diabetes. This was also good to hear (of course it’s needed!) – it was great news to hear earlier this week that we will see the impact of SGLT inhibitors in type 1 diabetes on cardioprotection when sub-group data emerges in heart failure studies for empa (apparently there are lots with type 1 in those studies).

  • Congratulations to Dr. Lora Heisler (University of Aberdeen) for winning this year’s Outstanding Scientific Achievement Award. Her brilliant award lecture reviewed her neuroscientific work into the effects of lorcaserin (Arena/Eisai’s obesity drug Belviq) and D-fenfluramine in the brain. Dr. Heisler has had an impressive career, breaking ground all the way. When she first began research, the notion of drugs that act in the brain to improve metabolism was absolutely (!) unheard of. Funding for research into this topic was hard to come by. Fast forward to today, and GLP-1s (which have targets in both the brain and the gut) are all the rage for the treatment of diabetes/obesity.

It’s always bittersweet when ADA’s Sci Sessions come to an end. We’re sad to say goodbye to our colleagues in Orlando, though we’re excited to return back home to San Francisco, and to bring you highlights from the fifth and final day. Fittingly, we’ve got five highlights! In rapid-fire fashion…

1. The writing committee for ADA/EASD’s new consensus statement, chaired by none other than Dr. John Buse and Prof. Melanie Davies, presented their draft recommendations. These include GLP-1s for ASCVD, SGLT-2s for heart failure, metformin remains first-line, and more…

2. Dr. Julio Rosenstock shared high-level topline data from Lilly/BI’s EASE program (SGLT-2 Jardiance in type 1)…

3. Dr. David Cherney argued for a CVOT enrolling people with type 1 diabetes…

4. Dr. Hiddo Heerspink described how common it is for a DKD drug to have responders vs. non-responders…

5. Dr. Lora Heisler gave an award lecture on diabetes drugs with therapeutic targets in the brain!

Don’t forget, you can find ALL our highlights from ADA 2018 on our Resource Hub: Our team is already cracking on the full report. Happy reading.

Diabetes Therapy Highlights

In a packed early-morning session, the ADA/EASD 2018 consensus committee outlined the major recommendations in its draft report on type 2 diabetes management. Perhaps most notably, the committee has endorsed SGLT-2s/GLP-1s for patients with CV disease, and promotes GLP-1 agonists as a first injectable therapy ahead of insulin. A webcast of these presentations will be live at this link in the next couple of days, and the draft report will be open for public comment (to be sent to until Monday, July 2 at 11:59 PM EDT; we highly recommend watching the full presentation before commenting as this report presents only our topline take. Overall, we see these guidelines as a much-needed step forward; the last ADA/EASD position statement was published in 2015, and we’ve seen multiple landmark trials since then (EMPA-REG, LEADER, CANVAS, DEVOTE) that should be influencing clinical care. We encourage all to watch the full presentation here when it goes live, and for anyone with positive or constructive feedback to send comments. As Dr. John Buse explained, people who disagree are often more likely to write-in, but the committee needs an accurate picture of how the field perceives this draft. Committee co-chairs Dr. Buse and Prof. Melanie Davies (what a duo!) led Drs. David D’Alessio, Judith Fradkin, Walter Kernan, Chantal Mathieu, Geltrude Mingrone, Peter Rossing, Apostolos Tapas, and Deborah Wexler in updating the ADA/EASD position statement. Notably, our report on the 2015 update includes “cardiovascular” only twice! Below, we outline the most notable components of the new draft consensus, with the caveat that this is not the final version.

  • GLP-1s and SGLT-2s will be recommended as second-line therapy in any patient with established CV disease, with additional specificity based on whether atherosclerosis or heart failure is the primary concern. The speakers mentioned repeatedly that metformin (+ lifestyle) will still be the first-line, foundational intervention for type 2 diabetes. If ASCVD is the predominant pathophysiology, a GLP-1 agonist with CV benefit should be prescribed; the committee members have even established an order of preference, with liraglutide (Novo Nordisk’s Victoza) ranked above semaglutide (Novo Nordisk’s Ozempic) ranked above long-acting exenatide (AZ’s Bydureon). If the patient isn’t tolerating GLP-1, then the committee suggests an SGLT-2 inhibitor with proven CV benefit, provided eGFR is high enough (≥45 ml/min/1.73 m2 in the US). The committee considers empagliflozin (Lilly/BI’s Jardiance) preferable to canagliflozin (J&J’s dapagliflozin) – we aren’t sure if this will persist given data of the last week. If A1c remains above target, the draft stipulates that HCPs should add the other class if it has not been attempted before (i.e. an SGLT-2 on top of GLP-1, or GLP-1 on top of an SGLT-2), a DPP-4 inhibitor (if not on GLP-1), basal insulin, a TZD, or an SU. That said, if heart failure is the predominant pathophysiology, the committee recommends against TZDs and saxagliptin (AZ’s DPP-4 inhibitor Onglyza).

    • The SGLT-2 inhibitor ranking makes sense to us, since canagliflozin showed an amputation signal in CANVAS that had yet to be sorted out when the guidelines were put together. AZ also manufacturers an SGLT-2 inhibitor, Farxiga (dapagliflozin). As Prof. Philip Home alluded to during Q&A, the DECLARE CVOT for dapagliflozin will likely be reporting later this year. Does this mean we’ll have to wait another three years for dapagliflozin to be incorporated into this consensus statement (provided DECLARE is a positive CVOT)? Dr. Will Cefalu described how the ADA Standards of Care was just this year made into a “living document” that can be edited more than once a year. He explained that this is easier to do with one organization vs. two, but committee members seemed in agreement that updates to the document may be needed sooner rather than later, as our insight on cardioprotection is rapidly changing. We were surprised to see liraglutide preferred over semaglutide by the writing committee, given how potent semaglutide is (some even call it best-in-class). That said, in terms of CV outcomes data, SUSTAIN 6 was a shorter, smaller trial than LEADER.

    • During Q&A, Dr. Davies clarified that this stands as a hard recommendation for any patient not at target (presumably for now defined by A1c). For patients with established CV disease who are meeting glycemic goals, however, the draft statement recommends that consideration be given to either switching or adding a cardioprotective diabetes therapy (i.e. a softer recommendation). And importantly, in this committee’s proposed framework, step no. 1 for a healthcare provider is assessing CV disease history and CV disease risk in their patients with type 2 diabetes – it’s so critical that CV risk reduction strategies are a central piece of diabetes care, even if people don’t have access to GLP-1s and SGLT-2s or don’t take them for any other reason.

    • Dr. Wexler noted certain limitations in the evidence used to generate these recommendations: For one, there’s no evidence of CV benefit in patients without prior events, so these trials really only apply to ~20% of the total population with type 2 diabetes. That doesn’t necessarily mean the benefits don’t exist in primary prevention, only that this hasn’t been adequately studied. One CANVAS post-hoc was cited a couple times: Originally presented at AHA 2017, this analysis found significant heart failure benefit with canagliflozin in the primary prevention cohort, but no significant difference on MACE events. Additionally, there is no CVOT data supporting the use of GLP-1s and SGLT-2s in combination. The relative preference of agents is based on a single trial for each, and it’s difficult to discern whether within-class differences are due to trial design or the molecules themselves. Finally, heart failure was a secondary endpoint in both CANVAS and EMPA-REG, meaning these events were not adjudicated quite as rigorously.

  • GLP-1 agonists are recommended as first injectable therapy before insulin when a patient is not at target with oral agents. While injectable therapy is still framed as an “escalation” of treatment – which one attendee questioned as semantically harmful (similar to how it’s dangerous to position insulin as a later or last resort) – we are so excited to see GLP-1 gain such a prominent spot. Dr. D’Alessio explained that the vast majority of patients will probably still prefer oral medications, and he clarified that clinical characteristics should still be considered when selecting an injectable. For example, the committee prefers basal insulin in the case of harmful weight loss or catabolism (indicating serious insulin insufficiency). Dr. D’Alessio emphasized that weight is a concern for almost every patient with type 2 diabetes; while some fixate on it more than others, GLP-1 agonists clearly have an edge over basal insulin, as the former gives weight loss while the latter spurs weight gain. Further, Dr. D’Alessio highlighted the superiority of GLP-1 agonists over basal insulin on hypoglycemia, another very important outcome for patient safety and quality of life. If a patient is already on a GLP-1 agonist and basal insulin is needed for further glucose-lowering, Prof. Mathieu made a plug for fixed-ratio basal insulin/GLP-1 combinations (Sanofi’s Soliqua and Novo Nordisk’s Xultophy). She explained that basal insulin can be intensified with a GLP-1, SGLT-2, or prandial insulin, as outlined in the draft statement. If a patient is on both GLP-1 and basal insulin and isn’t meeting A1c goals, prandial insulin should be added in a stepwise manner.

    • The committee did draw distinctions between specific insulins in figures and tables, but these were not reviewed in this session. We’d venture a guess (since these thought leaders certainly appreciate glycemic outcomes beyond A1c and hypoglycemia) that insulin degludec (Novo Nordisk’s Tresiba) will be preferred given the 40% risk reduction for severe hypoglycemia vs. glargine (Sanofi’s Lantus) in DEVOTE although it would be a big deal for them to recommend a next-generation insulin so we will wait to see.

  • Committee members did not find sufficiently robust evidence that initial combination therapy leads to long-term improvements in outcomes. “The evidence for dual therapy itself is weak and going beyond dual therapy is an evidence-free zone.” This was disappointing from our view given that so many patients are diagnosed with A1cs well about 7% and since there are multiple other therapeutic areas where combination therapy is a given – that said, we suspect they are just looking for more eve. Moreover, speakers acknowledged higher cost with polypharmacy, and in general established cost as an important variable in diabetes treatment decisions. We note that AACE currently recommends initial dual therapy for any patient with A1c ≥7.5% and initial triple therapy for any patient with A1c >9% + symptoms. ADA has historically been more conservative than AACE in its medication guidelines, and the 2018 Standards of Care suggest dual therapy for anyone with baseline A1c >9%. This ADA/EASD-commissioned committee only recommends initial combination therapy in people who are at least 1.5% above their personal target, which brings us to another key point…

  • This report aims to provide recommendations on how best to reach glycemic goals. It does NOT aim to pre-specify those goals for the entire type 2 diabetes population. This is an incredible nod to personalization of targets, in our view, and it could be a more effective strategy in improving real-world outcomes instead of arguing with ACP or any other professional organization about <8% vs. <7% vs. <6.5% and what will work best for the broad patient population.

  • While the jury may still be out on initial combination therapy, we would like to see influential professional organizations drive a shift toward earlier, faster intensification of treatment. Patients everywhere spend far too much time in hyperglycemia and out-of-range, not meeting goals on metformin. Could a more aggressive shift in algorithms make a dent in clinical inertia? What could this consensus statement say that would abolish, once and for all, the treat-to-fail paradigm? We have our wheels turning for what comments we might submit to this committee, and this lack of recommendation for earlier combo therapy is certainly on our minds.

  • GLP-1s and SGLT-2 are now considered safe for patients with CKD and high CV risk. The draft notes the safety of both classes in this set of patients (with dose reductions required for some medications), as well as the renal and CV benefits that some molecules have demonstrated even at a low eGFR. A CANVAS post-hoc presented at this very meeting (by Dr. Dick de Zeeuw) illustrates this latter point (canagliflozin gave cardioprotection to people with low baseline eGFR, even though A1c-lowering was attenuated for this subgroup). Because SGLT-2 CVOTs have found no excess adverse events in those with an eGFR <60 ml/min/1.73 m2 and have enrolled people down to 30 ml/min/1.73 m2, the committee feels safe recommending them at reduced renal function. GLP-1 agonists are specifically not recommended in end-stage renal disease due to limited experience and increased GI side-effects at lower eGFR.

    • Despite this ringing endorsement for safety, the committee is not (yet?) currently recommending SGLT-2s (or GLP-1s) for renal protection. We point to strong renal benefits seen in CANVAS and EMPA-REG OUTCOME. The panel explained that ongoing renal outcomes trials will offer better evidence on this endpoint in the next couple of years. During Q&A, Dr. Silvio Inzucchi questioned whether the evidence for renal protection was not as strong as the evidence for heart failure, and the committee didn’t disagree. We do find it a somewhat curious choice to make explicit recommendations based on heart failure but not on renal outcomes given equally compelling evidence, though the heart failure benefit to the SGLT-2 class has been corroborated by a number of real-world studies (CVD-REAL, EASEL, OBSERVE-4D) in a way that the renal benefit has not (yet). Might renal protection be the new cardioprotection in the next iteration of this consensus statement?

  • We noticed an increased focus on patient-centered care and shared decision-making from this writing committee. Dr. Fradkin (acting outside of her capacity at NIH) explained the basic balancing of risks and benefits around personalized treatment goals, taking into account issues of hypoglycemia risk, life expectancy, regimen complexity, disease duration, and CV status in answering the question: How much does a patient stand to benefit from stringent glucose control? From there, treatment should be selected with patient preferences at the very center, also taking into account issues of cost, efficacy, side-effects, and risks. Dr. Fradkin also highlighted the role of behavioral interventions and metabolic surgery, in a presentation otherwise very focused on pharmacotherapy.

  • The draft statement offers sample algorithms. If cost is a major issue, SUs and TZDs are recommended before adding insulin (NPH preferred) or a DPP-4 or SGLT-2 with the lowest acquisition expense. If weight is a significant concern in those without CV disease, SGLT-2s and GLP-1s are recommended after metformin, followed by whichever of the two the patient has not tried, followed by DPP-4 inhibitors (SUs and basal insulin come with warnings). Another algorithm is focused on glucose control while minimizing hypoglycemia, which offers four separate pathways and discourages the use of insulin and SUs.

  • Dr. John Buse brought this session to a close by outlining knowledge gaps and lingering questions. We were particularly interested in his comments on metformin as first-line. “Is metformin’s role as foundational therapy evidence-based, or just a quirk of history? If SGLT-2s were the first agents that arrived, is it possible that they would be the foundational therapy, and we’d be wondering if metformin should move up in the algorithm?” Indeed, we’ve attended several debates on this topic at recent conferences, and we do think we’ll see a day in the not-too-far future when metformin is no longer first-line. This is especially likely, in our opinion, once DPP-4s, GLP-1s, and SGLT-2s go generic. Dr. D’Alessio explained that metformin first was debated at length within the writing committee as well, but the composite of efficacy, safety, tolerability, and cost led to this decision to maintain status quo.

2. Dr. Rosenstock On EASE Topline Results (Lilly/BI’s SGLT-2 Empagliflozin in T1D): Can Low 2.5 mg Dose Offer Equivalent Efficacy w/o DKA Risk? Possibility for Type 1 CVOT?

Dr. Julio Rosenstock shared the first phase 3 findings on Lilly/BI’s SGLT-2 inhibitor Jardiance (empagliflozin) in type 1 diabetes; these were very high-level topline results (no specific numbers), and a full results presentation is expected at EASD (on October 4). The EASE program (n=1,680) randomized patients with type 1 diabetes to empagliflozin or to placebo. EASE-2 (n=720) and EASE-3 (n=960) both met their primary endpoint of superior A1c reductions with the SGLT-2 vs. placebo after 26 weeks, and very notably, the 2.5 mg empagliflozin dose used in EASE-3 demonstrated significant glycemic efficacy (and weight loss efficacy) without a significant spike in DKA. Jardiance for type 2 diabetes is available in 10 mg or 25 mg pills, and Dr. Rosenstock described the background to evaluating an even lower dose in type 1: 2.5 mg tablets were used in the phase 2 dose-ranging EASE-1 study, and showed remarkably similar effects to the 10 mg tablets. He speculated that 2.5 mg empagliflozin could have hit the “sweet spot” of adjunct oral treatment for type 1 diabetes. We can’t assess this hypothesis until we actually see the EASE data, and we look very forward to EASD for this. Recall that DEPICT-1 showed no imbalance in DKA between AZ’s Farxiga (dapagliflozin) and placebo at 24 weeks, but this serious safety risk appeared in follow-up out to one year. Sanofi/Lexicon’s SGLT-1/2 dual inhibitor sotagliflozin has been linked to elevated DKA risk in all phase 3 inTandem trials so far. On the other hand, DKA does seem to be dose-dependent to some extent, based on our understanding of all DEPICT and inTandem results presented or published to-date. We’re more intrigued than anything else about how 2.5 mg empagliflozin performed in EASE-3, and it would of course be a big win for patients if this formulation uniquely did not show any DKA risk signal (this would seem unlikely but we have not heard KOLs discuss it at length). We’re in wait-and-see mode. Lilly/BI also issued a factual press release about this topline data, which you’ll find here.

EASE-2 + EASE-3 Summary of Results

  • Dr. Rosenstock reviewed study design for both phase 3 EASE trials. EASE-2 started with six weeks of insulin optimization, followed by two weeks of placebo run-in. Type 1s were then randomized to empagliflozin 10 mg, empagliflozin 25 mg, or placebo for 52 weeks total, though the primary endpoint was read at 26 weeks. EASE-3 was similar in design except that there was a fourth treatment arm – 2.5 mg empagliflozin – and there was no trial extension after week 26. It’s important to note that EASE-3 was a relatively shorter study, especially if the conclusion is that this lower dose doesn’t confer excess DKA risk (again, DKA risk did not appear in the first phase of DEPICT-1 but became apparent after a full year of follow-up). Both EASE studies collected three weeks of follow-up data after patients stopped their assigned therapy. Baseline A1c was a mean 8.1% in EASE-2 and 8.2% in EASE-3, while baseline BMI was 29 kg/m2 and 28 kg/m2, respectively. Mean duration of type 1 diabetes was 23 years in EASE-2 and 21 years in EASE-3; EASE-2 included 41% pump users (the rest MDI) and EASE-3 included 66% pump users (the rest MDI).

  • Despite the potential promise of 2.5 mg empagliflozin for type 1, Dr. Rosenstock emphasized that DKA is a real and serious concern when giving therapy that is adjunct to insulin. He shared a conservative perspective on DKA risk management, arguing that all type 1 patients taking an SGLT inhibitor should regularly monitor blood ketones – urine strips are not enough, in Dr. Rosenstock’s view. “I don’t think people with type 1 should be using these drugs without having a blood ketone monitoring tool,” he established, also noting that DKA is preventable if it’s detected early (hence the push for more aggressive and diligent checking). We spent Friday morning in a room full of 30 thought leaders debating questions like this (how often type 1s on an oral adjunct should check ketones, which method they should use, whether basal or bolus insulin should be reduced, etc.), at a gathering organized by Professor Thomas Danne and ATTD. Some shared commentary in line with Dr. Rosenstock’s remarks, while others opt for a less stringent DKA risk management protocol in their clinical practice (currently, SGLTs are only prescribed off-label to type 1s). For example, Prof. Chantal Mathieu suggested that asking a patient to check ketones every morning imposes an undue burden and is unsustainable (meaning people will do it for a few weeks, and then will taper off or stop); she reinforced that in clinical studies, most DKA events occurred randomly during the day in the absence of elevated morning ketones. In Prof. Mathieu’s view, it’s more valuable for patients to learn to check ketones when they feel unwell. This area still has some ambiguity but we believe experts will reach consensus on this front. We’re leaning toward a more conservative approach to DKA risk management as well, because we want to make sure these products are approved for type 1 diabetes and are used safely in real-world patients, though we’re refining our thinking as we continue to absorb insights. As Dr. John Buse put it during the Friday consensus meeting, “I’d rather that some people just not get SGLT inhibitors based on our recommendations, rather than see an epidemic of DKA after these drugs are approved. I’d rather the 20% of patients who could really benefit get this drug.” From our view, there is incredible consensus already that “patient selection” is incredibly important – which patients specifically should be left out is also evolving.

  • As for the competitive landscape, AZ has submitted SGLT-2 inhibitor Farxiga (dapagliflozin) for a type 1 indication in Europe and Japan. The company plans to follow-up with an FDA submission in the second half of this year. Sanofi/Lexicon have filed their SGLT-1/2 dual inhibitor sotagliflozin with both FDA and EMA, and recently announced the intended brand name “Zynquista.” In all likelihood, FDA will convene an Advisory Committee meeting to discuss DKA and the risk/benefit profile of SGLT inhibitors in type 1 diabetes. We suspect this will take place in DC around January, and perhaps FDA will consider having one meeting for both AZ’s and Sanofi/Lexicon’s molecules?

  • During Q&A, Dr. Rosenstock pointed to the “opportunity” for a post-marketing CVOT in type 1 diabetes with one of these drugs. Given the high cost of running a CVOT, we wonder if manufacturers would ever consider taking this on themselves or whether it would have to be FDA-mandated. We agree it would be incredibly valuable for patients and for the diabetes field. See directly below for Dr. Rosenstock’s comments. As we understand it, BI/Lilly may have data on this front after the heart failure studies appear since the sub-group element has a high number of both type 1 and type 2 patients - we are eager to hear more on this front.

Select Questions and Answers:

Dr. William Cefalu (ADA): Were there any particular CGM parameters you can share?

Dr. Rosenstock: There was a significant increase in time-in-range. And there was nice data on glucose variability.

Dr. Cefalu: I guess we’ll hear about it at EASD.

Comment: I would have thought that the renal protection of SGLT-2 inhibitors would be particularly relevant for the type 1 diabetes population, where mortality and reduced life expectancy is particularly seen in those with kidney disease.

Dr. Rosenstock: I totally agree. We’ve done a lot of CVOTs in type 2 diabetes, with more than 200,000 patients involved. I do think there’s an opportunity to do a post-marketing outcomes study in type 1 diabetes with established CV disease, and I hope it’s a requirement.

3. Dr. Cherney Advocates for CV/Renal Outcomes Studies in Type 1 Diabetes; Offers Deep Dive on SGLT Mechanisms in T1 vs. T2

Dr. David Cherney gave the final talk of ADA 2018, on a topic that’s been a major theme of this entire conference: SGLT inhibitors in type 1 diabetes. He outlined compelling reasons to believe that this therapy class could be CV/renal protective for patients with type 1, concluding that there’s an “urgent need” for cardio/renal outcomes trials in type 1 diabetes. While the mechanisms of SGLT inhibition should play out in type 1 diabetes much like they do in type 2 (Dr. Cherney highlighted natriuresis and hypoxia in particular), he acknowledged that we have no “hard evidence for end organ protection.” While Lilly/BI’s EMPA-KIDNEY outcomes trial of SGLT-2 inhibitor Jardiance (empagliflozin) in CKD will include a subpopulation with type 1 diabetes, Dr. Cherney speculated that this will be a small minority of patients, and that sub-analyses won’t be sufficiently powered. Lilly/BI representatives at Friday’s SGLT/DKA consensus meeting noted that the heart failure program around Jardiance (EMPEROR HF) will also enroll some participants with type 1 diabetes, and we imagine Dr. Cherney would have the same reservations about sub-analyses conducted here. In other words, we can’t rely on sub-analyses to yield insights on how SGLT inhibitors affect diabetes complications for people with type 1. (To our understanding, AZ’s parallel Dapa-CKD and Dapa-HF outcomes trials exclude patients with type 1 diabetes.) A recent paradigm shift in type 2 diabetes care has positioned CV risk reduction as a central goal of any treatment plan; the most noticeable change in type 1 diabetes care of late has been growing emphasis on glycemic outcomes beyond-A1c, but to be sure, CV and renal risk reduction are also of key interest for this patient population. Dr. Anne Peters has explained that many of her older patients with type 1 ask for an off-label SGLT-2 inhibitor because they’ve heard about the CV benefits in type 2 diabetes, and they want the same for themselves.

  • Dr. Cherney listed several barriers to a type 1 diabetes CVOT as well, giving balance to his presentation. DKA is a small but significant risk associated with SGLT inhibitors, and he implied that investigators may not want to pass out this medicine to thousands of patients until better safety protocol are in place. As a reminder, Dr. Julio Rosenstock’s suggestion earlier in this same symposium was that AZ, Sanofi/Lexicon, or Lilly/BI conduct a post-market CVOT for their molecule in type 1 diabetes; presumably, safety procedures and risk management strategies will be more solid by the time these products reach the market (hopefully starting in mid-2019). We expect more counsel to be sought on this front. Dr. Cherney continued by mentioning the need for creative funding sources (perhaps a public-private partnership) and the need to enrich the patient population with established CV disease or DKD at baseline. He spoke to lessons learned from type 2 diabetes CVOTs in terms of enriching the study population to have an adequately powered trial. He also asked the question, “which outcomes?”, and we’d hope that such a trial would be very thoughtfully designed to measure the endpoints most meaningful in type 1 diabetes. We don’t think it’s too early to start talking about a type 1 CVOT, because we’re optimistic that SGLTs will eventually be approved for this indication, and regardless, we imagine patients will continue to take them off-label (so FDA may as well regulate safety, and research dollars should be invested in elucidating micro and macrovascular effects).

4. Dr. Heerspink On Responder/Non-Responder Variation in DKD Treatment; Defends Albuminuria as an Endpoint

In a standing-room-only session, clinical pharmacologist Professor Hiddo Lambers Heerspink added nuance to our understanding of renal benefits with different diabetes drugs by pointing out heterogeneity in albuminuria-lowering response. A recent sub-analysis of the MARLINA-T2D trial illustrates this well. Presented at ADA 2016, MARLINA-T2D found numerically higher but not statistically significant renal benefits for linagliptin (Lilly/BI’s DPP-4 inhibitor Tradjenta) in people with DKD (n=630). A deeper dive into the data reveals that a majority of patients in the study fell into two categories of albuminuria-lowering response to linagliptin: 44% experienced an impressive >20% reduction in albuminuria while 37% experienced an increase in albuminuria, indicating worsening of kidney function. Very few participants experienced anything in between, whether it be more modest 10%-20% or <10% reductions in albuminuria, suggesting a responder/non-responder effect with this DPP-4 inhibitor when it comes to the kidney. Dr. Heerspink hinted that there is more MARLINA-T2D data to come at EASD 2018 in October, and that a similar heterogeneity in effect also exists for albuminuria response to liraglutide (Novo Nordisk’s GLP-1 agonist Victoza). He alluded to a not-too-distant future of personalized therapy for DKD tailored to a person’s responder/non-responder status. He phrased this as a “one fit for everyone” approach, in contrast to the current “one size fits all” paradigm for DKD. We previously heard Dr. Heerspink speak on this topic at WCTD 2017, where he described how this responder/non-responder effect has precluded nephropathy research and drug development for kidney disease in diabetes. Indeed, no new DKD indications have been approved since irbesartan and losartan in 2000, nearly two decades ago.

  • Also on the subject of therapy personalization, Dr. Heerspink pushed back against the widespread recommendation that SGLT-2 inhibitors be discontinued in people with CKD. The logic of this recommendation is that SGLT-2s (which depend on the kidneys for their mechanism of action) will lose their glucose-lowering ability as eGFR falls, which isn’t untrue. However, Dr. Heerspink contended that this is precisely the patient population that has the most to gain from SGLT-2 inhibitor therapy because this class is associated with CV/renal protection. Dr. David Fitchett shared a similar opinion at ESC 2017, and we look forward to dedicated outcomes studies of SGLT-2s in heart failure and CKD to provide hard evidence on this question.

  • Dr. Heerspink argued against the idea that albuminuria is a “soft” and uninformative renal endpoint, noting that meta-analyses have shown a reliable connection between albuminuria and more widely-accepted “hard” renal endpoints like eGFR and progression to end-stage renal disease. Specifically, a 25%-30% reduction in albuminuria strongly predicts that a drug will also decrease the risk of progression to end-stage renal disease. Because these latter endpoints can only be assessed in very long-term studies, Dr. Heerspink suggested that albuminuria is arguably an even more useful indicator of the nephron-protective potential of a drug because it can be evaluated in trials of all kinds. He explained how this issue was discussed at a recent National Kidney Foundation FDA/EMA workshop, and the EMA now officially recognizes albuminuria as a valid surrogate renal outcome for primary prevention of CKD alongside rate of eGFR decline.

5. Outstanding Scientific Achievement Awardee Dr. Lora Heisler Discusses Therapeutic Targets for T2D in the Brain

University of Aberdeen’s Dr. Lora Heisler delivered this year’s Outstanding Scientific Achievement Award Lecture, positioning the brain as the next frontier of type 2 diabetes treatment. Dr. Heisler’s pioneering neuroscience work has focused on elucidating the mechanism of obesity drugs D-fenfluramine (now discontinued due to CV side-effects) and lorcaserin (Arena/Eisai’s Belviq). Both agents mimic the action of serotonin on the 5HT-2C receptor in the brain, which then stimulates the release of α-MSH, which in turn binds to MC4R receptors, thereby activating the appetite-suppressing pro-opiomelanocortin pathway. Dr. Heisler’s preclinical research found that this mechanism is also associated with glycemic benefits independent of weight loss. We’ve previously heard from obesity thought leader Dr. Ken Fujioka that lorcaserin can produce a rapid beneficial effect on blood sugar before weight loss sets in, and this fits in with Dr. Heisler’s remarks. It’s unclear whether lorcaserin could be reformulated as a dedicated diabetes drug (we’re not aware of any companies working on this), but the possibility is intriguing. Dr. Heisler contextualized that when her research career first began, the notion of drugs that act in the brain to improve metabolism was completely unheard of, and funding for research into this topic was hard to come by. Fast forward to today, and GLP-1 agonists (which have targets in both the brain and the gut) are all the rage for the treatment of diabetes and obesity, and the obesity drug competitive landscape abounds with neural-acting candidates like FGF-21 analogs, MetAP2 inhibitors, and MC4R agonists (which, like lorcaserin, act on the pro-opiomelanocortin pathway). We applaud Dr. Heisler for her incredible contributions to the science of diabetes and obesity, and for quite literally laying the foundation for the next-generation of therapies for metabolic disease.


-- by Ann Carracher, Abigail Dove, Payal Marathe, and Kelly Close