ENDO 2016 (98th Annual Meeting of the Endocrine Society)

April 1-4, 2016; Boston, MA; Days #1-2 Highlights – Draft

Executive Highlights

Greetings from Boston, where we are welcomed by a decent amount of springtime snowfall (!), as we bring you our first two days of coverage from the Endocrine Society’s 98th Annual Meeting (ENDO 2016). As 2016 marks the Endocrine Society’s centennial, this meeting has started with a bang, matching the milestone’s excitement. Day #1 opened with Dr. Ed Damiano’s big announcement to commercialize the Bionic Pancreas (read our in-depth report from Friday for a deep dive) and Dr. Doug Melton’s next steps on his team’s beta cell replacement therapy. In addition, we heard new data from AZ’s Farxiga (dapagliflozin) on DKA; Novo Nordisk debuted its fair share of data, including the first full phase 3 data for weekly injectable GLP-1 agonist semaglutide as well as phase 2 data on its oral GLP-1 program. The days’ corporate symposia also generated some noteworthy commentary around CVOTs. For our top ten highlights and four honorable mentions, please see below. Our preview details what’s coming up in the meeting’s second half; and if you missed our takeaways from ENDO’s preconference events, also take a look at our coverage of the Endocrine Fellow Series and the diabetes and obesity workshops.

1. During a captivating morning plenary, Dr. Ed Damiano (Boston University, MA) disclosed plans to commercialize the Bionic Pancreas via a new public benefit corporation, Beta Bionics. He also unveiled the third generation fully integrated iLet dual-chamber pump, which Beta Bionics plans to test in a four-site bridging study in 4Q16 and a pivotal study in 2Q17. See our detailed coverage from Friday for all the details.

2. Dr. Doug Melton (Harvard Stem Cell Institute, Cambridge, MA) discussed upcoming steps for his beta cell replacement therapy for type 1 diabetes: addressing small discrepancies compared to donor islets, producing other types of islet cells from stem cells, and protecting the cells from autoimmune attack.

3. The link between SGLT-2 inhibitors and euglycemic DKA in type 1 is solidifying, with new clinical data showing a 10% incidence of DKA and widespread increase in glucagon and ketones with AZ’s Farxiga (dapagliflozin) added to insulin + liraglutide.

4. The first full phase 3 data for Novo Nordisk’s weekly injectable GLP-1 agonist semaglutide showed substantial 1.5%-1.6% A1c reductions from a baseline of 8.1%, with better GI tolerability than we have seen in other semaglutide studies but with two blips on the safety radar.

5. The debut of phase 2 data on Novo Nordisk’s oral GLP-1 program revealed great efficacy (up to -1.9% A1c from a baseline of ~8%) but lots of nausea at max doses, supporting the phase 3 decision to go with lower doses that still have good efficacy but much better GI tolerability.

6. According to the esteemed Dr. Daniel Einhorn (Scripps Health, La Jolla, CA), it is hard to justify not preferentially prescribing Novo Nordisk’s Victoza (liraglutide) and Lilly/BI’s Jardiance (empagliflozin) in today’s post-LEADER and EMPA-REG era. We await news on whether the payers agree.

7. At an AZ-supported corporate symposium, the highly respected Dr. Zach Bloomgarden (Icahn School of Medicine at Mount Sinai, New York, NY) shared a personalized approach to prescribing SGLT-2 inhibitors and GLP-1 agonists in high CV risk patients, along with his optimistic take on the TZD pioglitazone (the latter has some baggage however).

8. We also saw a new analysis of the data from Lilly/BI’s Glyxambi (empagliflozin/linagliptin) phase 3 trials, which showed that the combination achieves consistent A1c changes irrespective of BMI, renal function, or gender.

9. Dr. Eric Vajda (Ligand Pharmaceuticals, La Jolla, CA) presented positive phase 1b data of Ligand’s glucagon receptor antagonist LGD-6972, demonstrating that the candidate substantially improves glycemic control and is well tolerated.

10. Our biggest takeaways from ENDO’s exhibit hall this year was the focus on concentrated insulins, most notably including Novo Nordisk’s Tresiba (insulin degludec) as well as Lilly’s new KwikPen for Humulin U500; obesity pharmacotherapies were also represented (to our slight surprise), with Eisai promoting Belviq (lorcaserin) and Novo Nordisk promoting Saxenda (liraglutide 3.0 mg).

Honorable Mention:

1. Dr. Steve Edelman (UCSD, San Diego, CA) delivered a passionate presentation on the benefits of CGM for the day-to-day management of type 1 diabetes, arguing that it “has been the single most important advance for type 1 diabetes since the discovery of insulin.” We certainly agree with that – to boot, particularly professional CGM can also bring great help to type 2 patients in reaching glycemic targets.

2. Pioneering obesity researcher Dr. Michael Schwartz (University of Washington, Seattle, WA) hypothesized that the pathogenesis of diet-induced obesity involves injury of hypothalamic neurons, leading to associated glial responses and structural alteration within the brain.

3. Novo Nordisk presented a poster on the LIRA-SWITCH trial, which demonstrated that Victoza (liraglutide 1.8 mg) provided superior A1c reductions compared to continued treatment with Merck’s Januvia (sitagliptin).

4. In this year’s presidential opening remarks, Endocrine Society President Dr. Lisa Fish (Hennepin County Medical Center, Minneapolis, MN) announced the upcoming launch of the society’s first open-access journal, the Journal of the Endocrine Society, as well as a new partnership with the Diabetes Education and Camping Association.

Top Ten Highlights

1. During a captivating and packed morning plenary, BU’s Dr. Ed Damiano disclosed plans to commercialize the Bionic Pancreas via a new public benefit corporation, Beta Bionics. He also unveiled the third generation fully integrated iLet dual-chamber pump, which Beta Bionics plans to test in a four-site bridging study in 4Q16 and a pivotal study in 2Q17. See Beta Bionics’ website here and our detailed coverage from Friday here. The updated iLet device is 17 mm thin (and ~33% smaller than the first-gen prototype), with a sleek aluminum exterior and glass capacitive touchscreen to improve the user experience. Beta Bionics holds the exclusive rights to the iLet device, as well as all additional Bionic Pancreas intellectual property previously owned by Boston University. Dr. Damiano demonstrated the easy iLet setup process for the audience, which simply requires body weight entry (no prior insulin or glucose data – this adaptation has long been a hallmark of the Bionic Pancreas algorithm and is unique in the field). The setup screen also includes functions to start and calibrate the CGM, place an infusion set, and load the insulin or glucagon cartridges. We appreciated the review of the intuitive Tidepool-developed user interface, which we first saw at FFL 2015 and heard about again at last November’s DiabetesMine event.

  • Dr. Damiano asserted that the Bionic Pancreas simultaneously solves the four greatest concerns in type 1 diabetes management: (i) it reduces mean blood glucose to lower the risk of long-term complications; (ii) it “profoundly curtails” hypoglycemia, and will likely eliminate severe hypoglycemia; (iii) it automates blood glucose control, unburdening people of the daily hassles of managing type 1; and (iv) it unburdens people with type 1 and their families of the emotional hardship and fear of hypoglycemia and long-term complications. Said Dr. Damiano. “A technology that could address any one of these concerns would be a breakthrough; a device that simultaneously solves all four is truly gamechanging and paradigm shifting.” We wholeheartedly agree and are thrilled to see progress to bring the system from the academic setting to patients – the automated insulin delivery competitive landscape gets more packed by the day, and it’s good to see many systems coming to compete with Medtronic’s MiniMed 670G (FDA submission before the end of June).
  • Dr. Damiano also reviewed previous clinical data from several prior Bionic Pancreas studies (The Bionic Pancreas Multicenter Study in 2014-2015, and the Summer Camp Studies in 2013 and 2014), remarking on the reduction in mean glucose and less time spent <60 mg/dl and >180 mg/dl in study participants vs. standard care. He emphasized the huge variability in daily insulin requirements in people with type 1 diabetes, noting that the Bionic Pancreas can adapt to these needs and “comply with the patient” rather than the other way around – “Do you want to follow your insulin, or do you want your insulin to follow you?” It is of course ludicrous to expect flat glycemia every day with the manual, static insulin dosing protocols used today in type 1 diabetes; we love that automated insulin delivery can adapt to that variability seamlessly with less patient burden. 

2. Dr. Doug Melton (Harvard Stem Cell Institute, Cambridge, MA) discussed upcoming steps for his stem cell-derived beta cell replacement therapy for type 1 diabetes. He first reviewed the research that went into developing the widely heralded differentiation protocol and the results showing functional capacity comparable to that of islet cells from cadaveric donors. He also noted that his team has applied the protocol to iPS cells in addition to embryonic stem cells and that those cells have also demonstrated glucose-responsive insulin secretion and the ability to reverse diabetes in a mouse model. Dr. Melton then listed several challenges his team aims to address in the near future. He explained that while his differentiated cells are functionally indistinguishable from donor islet cells when transplanted into animal models, there are still some small discrepancies in glucose-responsive insulin secretion, particularly at lower glucose levels, when the cells are analyzed individually ex vivo. As mentioned at last year’s ADA, his lab is also working on modifying the differentiation protocol to produce complete (or near-complete) islets with multiple cell types. To that end, they have done RNA sequencing on every cell from four donor islets to get a complete picture of the gene expression profile they are aiming for. They have successfully modified the differentiation protocol to produce alpha and delta cells in addition to beta cells (albeit less efficiently) and are working on producing cells that secrete ghrelin and pancreatic polypeptide. In addition to these incremental changes, Dr. Melton acknowledged that the big challenge will be finding a way to protect the cells from immune attack, either through encapsulation or biological protection. His group has already made significant progress toward this goal, recently publishing positive preclinical results with the cells using an alginate microencapsulation device developed by Dr. Dan Anderson (MIT, Cambridge, MA).

3. Valuable new data demonstrated a ~10% incidence of ketoacidosis with SGLT-2 inhibitors (specifically AZ’s Farxiga [dapagliflozin]) in type 1 diabetes, plus a marked increase in ketones across the patient pool. The specter of euglycemic DKA has stimulated much discussion recently – including an AACE/ACE meeting late last year – but there have not been enough data to provide insight on the mechanism and magnitude of the risk. Here at ENDO, Dr. Husam Ghanim (University of Buffalo, Buffalo, NY) presented the full results of a study that we got an interim look at during the AACE/ACE meeting. The study took 30 patients on insulin + liraglutide at baseline and randomized them to add dapagliflozin or placebo. The two DKA cases occurred immediately after the dose increase from 5 mg to 10 mg dapagliflozin, with low final insulin doses (≤ 0.5 U/kg) identified as the primary risk predictor. More broadly, there were statistically significant increases in plasma glucagon, free fatty acids, and ketones across the dapagliflozin group – the increase in certain ketones was more than three-fold, and enough (according to Dr. Ghanim) to take many patients close to a threshold for ketosis. If anything, the concomitant GLP-1 agonist therapy would be expected to be protective by decreasing glucagon levels, and the fact that increases in glucagon, ketones, and DKA cases were still seen is not reassuring. Still, we hope that protocols can be designed and disseminated to mitigate the risk of DKA (i.e.: by limiting insulin dose reduction to <20% and keeping the final dose above 0.5 U/kg), as even this study confirmed previously reported A1c and weight benefits with SGLT2 inhibitors. Recent work from dQ&A (contact Richard.wood@d-qa.com) showed that very few patients had access to a ketone meter or understood what appropriate ketone results should be. This is a very addressable problem in our view. 

  • A sneak peek at results from a study investigating liraglutide in type 1 demonstrated benefits on A1c and postprandial glucose that were pretty modest – see our Detailed Discussion & Commentary section below for more details.

4. The first full look at phase 3 data on Novo Nordisk’s once weekly injectable GLP-1 agonist semaglutide from SUSTAIN 1 painted a picture of strong efficacy and a good – though not spotless – safety and tolerability profile. As we first learned with the release of topline SUSTAIN 1 results last July, following 30 weeks of treatment, mean A1c fell 1.5% with semaglutide 0.5 mg and 1.6% with semaglutide 1.0 mg from a baseline of 8.1%, compared to essentially no change in the placebo group. Though impressive, we’ve actually seen even greater efficacy in other topline SUSTAIN program results. Body weight in SUSTAIN 1 fell around 3 kg (~7 lbs), placebo-adjusted. Impressively, both semaglutide doses helped three-fourths of patients achieve the composite goal of an A1c below 7% without weight gain or hypoglycemia. Relatively high rates of nausea have been an issue with earlier stages of the semaglutide clinical development program, but the 20-24% incidence of nausea seen in SUSTAIN 1 was more or less comparable with the rest of the GLP-1 agonist class, and most cases were mild. Given that SUSTAIN 1 enrolled patients on no glucose lowering therapy at baseline (not the most clinically relevant application for GLP-1 agonists), we’ll be looking even more closely at other SUSTAIN trials that add semaglutide to metformin and/or use an active comparator.

  • The only eyebrow-raising moment in the presentation was a four-to-zero imbalance in malignancies with semaglutide. However the malignancies were all of different types, and the 2-to-1 randomization clouds numerical comparisons of adverse events. Four patients on semaglutide developed gallstones as compared with none on placebo. The total number of events was too low to interpret these signals as definite, but we’ll be looking closely at malignancy and gallstone data in future SUSTAIN trial presentations.

5. As another big first, we finally got a look at full phase 2 results from Novo Nordisk’s oral semaglutide program – this was a truly expansive study (n=632) presented by Dr. Julio Rosenstock (Dallas Diabetes and Endocrine Center, Dallas, TX).  The trial covered a wide range of doses vs. placebo, had injectable weekly semaglutide as a comparison to provide clinical perspective, and even tested fast and slow titration for the high oral doses. And by high, we mean high – out of the range of oral doses tested (2.5 through 40 mg daily), the biggest is 280x more drug than the 1 mg weekly subcutaneous dose (this is due to low oral bioavailability). Efficacy was stellar for the 40 mg dose: an A1c drop of 1.9% (!) from a low baseline of ~8%, the same as was seen with the 1 mg injectable semaglutide group, plus a 51 mg/dl drop in FPG and 6 kg (~13 lb) placebo-adjusted weight loss after 26 weeks. However, we can see why Novo Nordisk appears to be moving forward with lower doses (3 – 14 mg daily) in phase 3, as there was a big jump between the 5 mg and 10 mg oral doses in the incidence of nausea (14% vs. 33%); treatment discontinuation leapt from 12% at 10 mg to 27% at 20 mg, the latter number probably being unacceptable. A slower up-titration regimen reduced discontinuations by nearly half, providing clues on what may work best in phase 3. Even a low 5 mg dose helped 81% of patients achieve an A1c below 7% (vs. 28% with placebo), so using lower doses, avoiding intolerable nausea, and positioning oral GLP-1 as an add-on to metformin seems like the way to go. Weekly injectable semaglutide could then be reserved for patients needing greater efficacy. Daily oral semaglutide has entered phase 3, which will include a CVOT which expected to start in July 2016.

6. Dr. Daniel Einhorn (Scripps Health, La Jolla, CA) shared a range of wisdom on the diabetes drug landscape, with Victoza (liraglutide), Jardiance (empagliflozin), Nesina (alogliptin), receiving the most positive mentions. Dr. Einhorn suggested that in today’s post-LEADER and EMPA-REG OUTCOME era, it would be “hard to justify not using these two classes of agents” (GLP-1 agonists and SGLT-2 inhibitors). Moreover, he specifically expressed preference for Victoza and Jardiance over competitors in their respective classes; he admitted that cardioprotection is most likely a class effect for both the GLP-1 agonists and SGLT-2 inhibitors, but that we cannot know for sure until we have data on the other drugs in those classes. He did, however, admit that formulary limitations are likely to be the primary determinant of choices between drugs in either class.

  • Earlier in the same Sanofi-sponsored morning symposium, Dr. Einhorn emphatically suggested that the TZD pioglitazone deserves to make a comeback – the rationale being: (i) clinical studies have largely cleared pioglitazone’s reputation of carrying a risk for cancer or major adverse cardiovascular events; (ii) the effect on fluid retention is minimal at low doses and in combination with GLP-1 agonists or SGLT-2 inhibitors; (iii) the drug is now an inexpensive generic, making it accessible to many patients. Further to that, the IRIS trial recently demonstrated a reduced risk of stroke and MI by 24% (p=0.007) and new incidence diabetes by 52% (p<0.001) compared with placebo in patients with insulin resistance, potentially revitalizing pioglitazone. He also gave Takeda’s DPP-4 inhibitor Nesina (alogliptin) mention for being extremely cheap, a marker of how competitive the DPP-4 inhibitor class has become.

7. At an AZ-supported corporate symposium, Dr. Zach Bloomgarden (Icahn School of Medicine at Mount Sinai, New York, NY) shared a personalized approach to prescribing SGLT-2 inhibitors and GLP-1 agonists in high CV risk patients. When asked what class to prescribe for a patient with coronary artery disease, Dr. Bloomgarden emphasized first examining all of the patient’s characteristics. Specifically, he noted that if the patient was at high risk for heart failure (i.e. had prior heart failure), SGLT-2 inhibitors may be the ideal way to go, given the impressive reductions in heart failure risk seen in EMPA-REG. On the other hand, Dr. Bloomgarden “preferred the idea of GLP-1 agonists” in individuals with obesity or good heart function. Overall, the symposium held onto a theme of the importance of individualizing therapy and understanding the heterogeneity of diabetes through its presentations and commentary. With regards to using this approach in high CV risk patients in the context of all the recent CVOT data, seeing the full results of the LEADER trial for Novo Nordisk’s Victoza (liraglutide) (which has positive topline results) will hopefully provide more color on what patients would benefit the most from which class.

  • On a somewhat similar note, Dr. Bloomgarden also suggested the potential of the TZD pioglitazone in lower doses. While the speakers (unsurprisingly) highlighted sulfonylureas as an inferior treatment choice, Dr. Bloomgarden noted that past studies of pioglitazone perhaps erred to the highest tolerable doses, which may have driven the well-known undesirable side effects such as weight gain and edema. However, he suggested that lowering the dose to 7.5 mg or 15 mg a day may maintain the drug’s benefits (i.e. risk reduction of stroke and MI as recently shown in the IRIS trial), while reducing the side effects. Dr. Derek LeRoith (Icahn School of Medicine at Mount Sinai, New York, NY) and Dr. James Gavin (Emory School of Medicine, Atlanta, GA) chimed in with support, expressing disappointment in “forgetting about insulin sensitization” and not using the “scientific concepts” the field “spent a lot of time developing.” With the positive results of the IRIS trial, pioglitazone has seemed to be picking up attention again – see our report on the findings for more of KOLs’ insights on TZDs.

8. A new analysis of data from Lilly/BI’s Glyxambi (empagliflozin/linagliptin) phase 3 trials showed that the combination achieves consistent A1c changes irrespective of BMI, renal function, or gender. As a reminder, the phase 3 trials of Glyxambi (presented at ADA 2014) found that at 24 weeks, the high-dose combination (empagliflozin 25 mg/linagliptin 5 mg) achieved A1c reductions of 1.1% and 1.2% (from baseline A1c of ~8%) as initial therapy and add-on to metformin, respectively. The analysis first cut the data by baseline BMI <25 kg/m2 vs. ≥25 kg/m2 – these results found consistent A1c reductions with the high-dose combination in both the initial therapy (A1c reductions of 1.13% [lower BMI] and 1.08% [higher BMI]) and add-on to metformin studies (A1c reductions of 1.21% [lower BMI] and 1.19% [higher BMI]). Similarly, there was a consistent effect regardless of baseline eGFR (60 to <90 ml/min/1.73 m2 vs. ≥90 ml/min/1.73 m2), with all arms achieving ~1% A1c reductions at the high dose of Glyxambi (1.08% and 1.06% in the low and high eGFRs, respectively, in the initial therapy study). As for gender, the effect also remained relatively consistent (in the add-on to metformin study, females and males achieved A1c reductions of 1.16% and 1.22%, respectively), except that females in the initial therapy study demonstrated a somewhat lower response compared to males (A1c reductions of 0.87% vs. 1.28%). Overall, these data supported the use of Glyxambi across a broad range of patients with type 2 diabetes, although it was noted in Q&A that it is important to take note of the lower response of females when prescribing therapy in this patient population. This first-in-class SGLT-2/DPP-4 inhibitor combination has certainly been a highly anticipated approach and the consistency found across the board in this analysis is encouraging in terms of incorporating this combination into future therapies, such as the triple-therapy tablet of empagliflozin, linagliptin, and metformin Dr. Ralph DeFronzo (University of Texas Health Science Center, San Antonio, TX) referred to at EASD – see our coverage of his presentation of Glyxambi’s 52-week results for more.

9. Dr. Eric Vajda (Ligand Pharmaceuticals, La Jolla, CA) presented positive phase 1b data of Ligand’s glucagon receptor antagonist LGD-6972, demonstrating that the candidate substantially improves glycemic control and is well tolerated. This randomized, double-blinded, placebo-controlled study (n=48) was a multiple ascending dose trial in normal healthy volunteers and individuals with type 2 diabetes. Over 14 days, normal healthy volunteers were administered a once-daily 15 mg dose of LGD-6972 while type 2 diabetes participants were randomized to either 5 mg, 10 mg, or 15 mg doses per day. The findings demonstrated that LGD-6972 reduced fasting plasma glucose in normal healthy volunteers (from ~97 mg/dl to ~86 mg/dl). In type 2 diabetes participants, the results found dose dependent decreases in fasting plasma glucose, with reductions of ~37 mg/dl, ~50 mg/dl, and ~53 mg/dl for the 5 mg, 10 mg, and 15 mg doses, respectively (from a baseline of ~160 mg/dl-200 mg/dl). From these findings, Dr. Vajda estimated an A1c reduction of ~1.5% at steady state with the highest dose. Notably, he also highlighted that these glucose reductions were observed in both fasting and postprandial states throughout the day. In addition, the study’s oral glucose tolerance test showed that LGD-6972 increased the insulin excursion and decreased glucagon excursion after an oral glucose load. Regarding the candidate’s safety and pharmacokinetic profile, the study had no serious adverse events (nausea and headache being the most common adverse events, mild to moderate in nature) and the candidate was well tolerated with no clinically significant changes in hematology, clinical chemistry, urinalysis, ECG, or vital signs. Dr. Vajda also pointed out that LGD-6972 was well absorbed with a long half-life (~50 hours) and that the drug’s steady state was achieved in a relatively short amount of time (within seven to 10 days). While this study’s duration is short, the data are promising for further clinical development. But as a small company, it may be challenging for Ligand to differentiate itself in the type 2 diabetes drug arena with the market’s increasing high bar, although LGD-6972 can be an appealing candidate to partner. For the company’s latest and more of our thoughts, please see our coverage of Ligand’s 4Q15 update.

10. Our biggest takeaways from ENDO’s exhibit hall this year was the focus on concentrated insulins, most notably including Novo Nordisk’s Tresiba (insulin degludec) as well as Lilly’s new KwikPen for Humulin U500. The star of the show was Tresiba, fresh off its US launch – not only did Novo Nordisk field a large front-and-center booth that was nearly entirely dedicated to Tresiba, but it also had Tresiba ads everywhere in the convention center. Interestingly, it was the more concentrated U200 formulation listed on most of the messaging. Also feeding into the conference-wide theme of more concentrated insulins, Lilly’s booth promoted the new KwikPen for Humulin U500, a more patient-friendly means of administration than the old vial and syringe. Humulin U500 has been a driver of Humulin franchise sales growth for a few years, with obesity and severe insulin resistance trending upwards, and the new pen should make the dose calculations less complicated. Lilly emphasized that Jardiance is leading the field in new patient starts in 2016. We saw Afrezza represented in Sanofi’s booth for perhaps the last time, following the end of the Sanofi-MannKind partnership.

  • We were also pleasantly surprised to see a decent amount of representation on the obesity pharmacotherapy front: Eisai had a substantially sized booth promoting its/Arena’s Belviq (lorcaserin) and Novo Nordisk had its separate booth focused on Saxenda (liraglutide 3.0 mg). Notably, this is the first time in a while we have seen Belviq on the exhibit floor – Arena/Eisai were not present at last year’s ENDO and its latest appearance to our knowledge was in the form of a small pop up booth at this past ADA. Similar to recent conferences, we did not see Orexigen/Takeda’s Contrave (naltrexone/bupropion extended-release) or Vivus’ Qsymia (phentermine/topiramate extended-release), indicative of the challenges these companies are facing.



Honorable Mention

1. “CGM has been the single most important advance for type 1 diabetes since the discovery of insulin.” Dr. Steve Edelman (UCSD, San Diego, CA) delivered a passionate presentation on the benefits of CGM in the day-to-day management of type 1 diabetes. He stressed that this positive view does not apply to blinded CGM, which he strongly believes is inappropriate in type 1 diabetes except in a research context. This was consistent with the view expressed by many participants at the recent AACE/ACE Consensus Conference on CGM, though a minority of attendees at that meeting argued that masked professional CGM has some value. It will be interesting to see how Abbott does with its blinded FreeStyle Libre Pro system in the US, which is expected to launch in mid-2016. Much of Dr. Edelman’s talk focused on the utility of real-time CGM to help patients make insulin-dosing decisions, particularly the significant adjustments patients make to bolus doses based on trend arrows (e.g., a BG of 220 mg/dl and two up arrows [take much more insulin] vs. two down arrows [take much less]). This work was first presented as a poster at ATTD 2014 and subsequently published in DT&T’s special CGM issue earlier this year. Dr. Edelman did not explicitly address adjunctive vs. non-adjunctive use of CGM, though he did briefly refer to a survey indicating 50% of patients would confirm a low CGM glucose value first before taking action, while the other 50% would treat right away. We assume more patients will begin using sensors to make treatment decisions once: (i) systems get insulin-dosing claims in the US (Dexcom expects it this year); (ii) accuracy and reliability continue to improve; and (iii) factory calibration arrives, putting fingersticks further in the background (Abbott’s FreeStyle Libre is “optimistically expected” by the end of 2016). Dr. Edelman repeatedly emphasized the benefits of CGM in eliminating the long-assumed tradeoff between tight control and severe hypoglycemia and took issue with the assumption that patients need to first be on a pump before beginning CGM – “I think that’s totally backwards.” We agree and look forward to data from the Dexcom’s Diamond study on this front at ADA 2016. Finally, Dr. Edelman half-jokingly noted that Medicare assumes his type 1 diabetes will disappear when he turns 65; while the bills to establish Medicare coverage of CGM are still in the works, we hope insulin-dosing claims and better cost-effectiveness data can convince CMS to reimburse the technology.

2. Pioneering obesity researcher Dr. Michael Schwartz (University of Washington, Seattle, WA) hypothesized that the pathogenesis of diet-induced obesity involves injury of hypothalamic neurons, leading to associated glial responses and structural alteration within the brain. He introduced obesity as a disorder of energy homeostasis, characterized by an increase in the defended level of body fat mass, delving into the latest research around Agrp and CGRP neurons. Specifically, Dr. Schwartz highlighted that Agrp neurons (located in the arcuate nucleus) are “crucial” in the body’s adaptive response to caloric deficit, as these neurons are activated by leptin deficiency. Thus according to Dr. Schwartz, if this adiposity negative feedback system involving leptin becomes impaired, the enhanced activation of Agrp neurons can be predicted to favor the defense of elevated body fat stores, leading to obesity. Recent work by Drs. Carlos Campos and Richard Palmiter, in collaboration with Dr. Schwartz’s lab, shows that Agrp neuron activation also inhibits neurons involved in satiety and meal termination, such as CGRP neurons (located in the external lateral parabrachial nucleus), whose activation in the rodent model can promote sustained anorexia. Their ongoing work defining interactions between CGRP neurons Agrp neurons exemplifies recent progress in the identification of neuron subsets and circuits involved in energy homeostasis and how dysfunction of these circuits might predispose to obesity. An example of this dysfunction is the neuroprotective response activated in microglia and astrocytes that occurs in response to such neuronal damage, and is evident in the hypothalamus of rodents made obese by consuming a high fat diet. According to Dr. Schwartz, this sequence of events can then result in structural alteration of key brain areas for energy homeostasis. We can imagine this sequence prompting a vicious cycle of damage within the brain, helping to explain why obesity is so challenging to treat. That said,  the more we know about the pathophysiology of  obesity, the greater our ability to identify effective approaches to treatment. For more on Dr. Schwartz’ insights on what’s next for obesity research and treatment, please see our recent interview with him.

  • A new “knockout round” session on Saturday also featured a focus on the brain’s role in obesity, with novel insights on neurotensin and the neural circuitry between leptin and ghrelin. Ms. Hilary Woodworth (Michigan State University, East Lansing, MI) discussed the interplay between dopamine neurons and the neuropeptide neurotensin, as she demonstrated that knocking out neurotensin-responsive dopamine neurons in the rodent model is associated with a phenotype that protects obesity. Specifically, these knockout mice remained slim despite consumption of a high-fat diet and engaged in significantly greater physical activity. In addition, Ms. Juliette Brown (Michigan State University, East Lansing, MI) highlighted the complexity of the neuronal circuitry involving the hormones leptin (which contributes to weight loss) and ghrelin (which promotes feeding). She focused on the lateral hypothalamic area within the rodent model’s brain, demonstrating that disruption of the ability to respond to leptin here not only disrupts the response to leptin, but the response of ghrelin as well, leading to obesity. While much of this research is in the early stage, the excitement around how the central nervous system drives obesity is clearly becoming a growing area of interest. On a separate note, we loved the fun and fast-paced dynamics of this “knockout round,” in which these young researchers competed on the basis of their concise and engaging presentation skills with only three minutes and one PowerPoint slide – an increasingly important skillset as attention spans decrease in this day and age!

3. Novo Nordisk presented a poster on the LIRA-SWITCH trial, which demonstrated that Victoza (liraglutide 1.8 mg) provided superior A1c reductions compared to continued treatment with Merck’s Januvia (sitagliptin). The trial’s findings showed that at 26 weeks, adults with type 2 diabetes who switched from Januvia (n=407) to Victoza achieved greater A1c reductions compared to those who continued Januvia treatment (n=204) (-1.1% vs. -0.5%; baseline A1c of 8.2%-83%). In addition, the Victoza arm experienced greater body weight reductions compared to the Januvia arm (-3.3 kg vs. -1.6 kg) (baseline weight of 89 kg-91 kg). Regarding safety, adverse events were more common with Victoza than with Januvia (69% vs. 57% of participants), with the most common adverse events being GI side effects in the Victoza arm. However overall, few serious adverse events were reported and no severe or confirmed nocturnal hypoglycemia was reported. While we would deem none of these findings as surprising, these head-to-head data further support the superior efficacy of the GLP-1 agonist class (but greater tolerability of DPP-4 inhibitors) and provide some additional guidance on potential next steps for patients uncontrolled on second-line therapy. See Novo Nordisk’s press release on the data for more.

4. In this year’s presidential opening remarks, Endocrine Society President Dr. Lisa Fish (Hennepin County Medical Center, Minneapolis, MN) announced the upcoming launch of the society’s first open-access journal, the Journal of the Endocrine Society. The journal is expected to launch in fall 2016 and will publish in basic, translational, and clinical endocrinology; past Society President Dr. Larry Jameson (Perelman School of Medicine, Philadelphia, PA) will serve as the journal’s first editor-in-chief. Additionally, Dr. Fish emphasized that the journal’s mission, through its open access, is to “bring greater equity to science and medicine through broad communication” – we applaud the Society for providing more venues to make the scientific field and the public smarter about endocrinology.

  • Dr. Fish also shared that EndoCares, the society’s “giving back” campaign, has partnered with the Diabetes Education and Camping Association. Not only would this work help young people with diabetes form valuable communities, the partnership also aims to bring together providers in new avenues. We love this interdisciplinary collaboration as it helps integrate different areas of expertise and brings the bigger patient perspective closer to the provider and scientific research at large.


Detailed Discussion and Commentary

Oral Presentations: Novel Treatment for Diabetes – Focusing on GLP-1 and SGLT2

Dapagliflozin Induces Ketosis in Patients with Type 1 Diabetes

Husam Ghanim, PhD (University of Buffalo, Buffalo, NY)

With a title like “Dapagliflozin Induces Ketosis in Patients with Type 1 Diabetes,” it was clear that the presenters of this oral presentation were not messing around. The study (n=30) provided one of the best datasets yet on the risk of DKA and mechanistic underpinning of ketosis seen with the use of SGLT-2 inhibitors in type 1 diabetes. This study was unique in that it administered an SGLT-2 inhibitor – AZ’s Farxiga (dapagliflozin) – in addition to not just insulin, but also the GLP-1 agonist liraglutide. A GLP-1 agonist would be expected to protect against DKA by blunting glucagon secretion, so the fact that both ketosis and diagnosed DKA occurred is particularly worrisome (unless the GLP-1 agonist therapy led to a large reduction in insulin dose, in which case it might not be protective). Before going to DKA, presenter Dr. Husam Ghanim noted that dapagliflozin had real benefits, like additional weight loss, no additional hypo, and what looked to be a modest A1c improvement. Dr. Ghanim’s group did a beautiful job demonstrating the mechanism at play behind the ketosis, by showing statistically significant increases in glucagon secretion (~35% increase), plasma free fatty acids (~70% increase), and hormone sensitive lipase (~25% increase) – the sequence of steps towards ketosis – along with the ketones acetoacetate (~50% increase) and beta-hydroxybutyrate (>200% increase). The risk appears real, though we still believe it is very manageable if appropriate patient and provider education occur optimally (contact Richard.wood@d-qa.com for more details on this front).

  • Dr. Ghanim shared details on both the cases of DKA that occurred in the trial. Both cases occurred shortly after the escalation from the 5 mg dapagliflozin dose to the 10 mg dose.
    • The first had euglycemic DKA, with a glucose concentration of <160 mg/dl. Her total daily insulin dose had fallen from 33 units to 26 units, which for her was only 0.35 U/kg.
    • The second case had hyperglycemic DKA. Her insulin dose had changed
  • A pressing action item for the diabetology community is to design and disseminate a protocol for mitigating the risk of DKA for type 1 diabetes patients taking SGLT-2 inhibitors. By all accounts there is a lot of consensus on the important points, namely: (i) starting and probably staying on the lowest dose available; (ii) capping the insulin dose reduction at 10%-20%; and (iii) keeping the absolute insulin dose from falling below 0.5 U/kg. Dr. Anne Peters (USC, Los Angeles, CA), who was one of the first to speak out about the euglycemic DKA story, shared her personal protocol for off-label use of SGLT-2 inhibitors in type 1 diabetes at last year’s EASD. Effectively educating stakeholders – especially patients and the hospitalists that may be the first line for confronting euglycemic DKA cases when they occur – will be critical to ensuring that SGLT-2 inhibitors have a future in type 1 diabetes care.   
  • Dr. Ghanim gave us a sneak peak at data from the original study investigating liraglutide in type 1 diabetes – these are to be published shortly in Diabetes Care. The benefit was apparent but modest, following the precedent of the larger LIRA-1 trial that was presented at ADA last year. The small size of the Buffalo study came through at a few points, such as when the liraglutide 1.2 mg group achieved a substantially larger placebo-adjusted A1c reduction from baseline (~0.5%, p<0.05) than the 1.8 mg group (~0.15%; p=NS). Liraglutide’s effect was more pronounced when it came to reductions in postprandial glucose AUC and glucagon, which were dose dependent and statistically significant across the board.

Questions and Answers

Q: In a recent study by Bob Henry with canagliflozin, they showed not only DKA, but also an increased risk of hypoglycemia. You said you did not see that here, but it looked like you had three episodes of hypo, with one being severe, with dapa, and just one with placebo. I do not think it is fair to say that there was not increase in hypoglycemia.

A: The difference in numbers was not statistically significant. The study also randomized two-to-one, so there were more patients on the drug than on placebo.

Efficacy and Safety of Once-Weekly Semaglutide Monotherapy Versus Placebo in Subjects with Type 2 Diabetes (SUSTAIN 1)

Christopher Sorli, MD (Billings Clinic Research Center, Billings, MT)

This presentation – in front of a packed house – featured the first results from the phase 3 program for Novo Nordisk’s once-weekly GLP-1 agonist candidate semaglutide. We already knew that the placebo-adjusted glucose lowering (~1.5% A1c reduction from a baseline of ~8%) and weight loss (~3 kg / ~7 lbs) results were fairly strong, as those topline results were released last year. As perhaps the best indicator of this product’s potential, both semaglutide doses helped three-fourths of patients achieve a final A1c below 7% without any weight gain or severe or BG-confirmed symptomatic hypoglycemia (vs. 19% of the placebo group). In addition, 59%-60% of the patients on semaglutide reached an A1c of 6.5% or lower. GLP-1 agonists, along with SGLT-2 inhibitors, tend to do especially well with these composite endpoints. We have seen unusually high rates of nausea elsewhere in the semaglutide clinical development program, but in SUSTAIN 1 nausea was reported in only 20%-24% of patients taking the drug, and the vast majority of cases were both mild and temporary. There were a couple imbalances in the safety analysis, namely a pair of four-to-zero imbalances vs. placebo with both malignancies and gallstones. Mitigating factors for the malignancies include: (i) there were two times as many patients in the trial taking semaglutide as placebo; and (ii) the cancer cases were not all of one type, raising the odds that the imbalance was due to chance.

  • As a test of semaglutide in treatment-naïve patients, SUSTAIN 1 is valuable for regulatory purposes but is not very clinically relevant given that GLP-1 agonists are not prescribed first-line. The double-blind trial randomized 388 type 2 diabetes patients on no baseline glucose-lowering therapy 1:1:1 to semaglutide 1.0mg, semaglutide 0.5 mg, and placebo for 30 weeks. The doses were escalated upwards over 4-8 weeks.
  • Notably, both semaglutide doses performed comparably when it came to glucose lowering efficacy, lowering A1c by 1.5% (0.5 mg) and 1.6% (1.0 mg) from a mean baseline of 8.1%. There was no mean A1c change from baseline in the placebo group. Presenter Dr. Sorli suggested during Q&A that in this trial, mean baseline A1c was low enough that even the lower semaglutide dose took patients to ~6.5%, and the higher dose had no more room to lower glucose further given that GLP-1 agonists only stimulate glucose lowering when glucose levels are high. We’d buy it, as we’ve seen greater magnitude A1c reductions out of semaglutide in other trials – including in the oral semaglutide abstract presented in this same session. Seven-point SMBG profiles showed clear reductions in both fasting glucose and postprandial glucose excursions.
    • Body weight fell by a placebo-adjusted 3.5 kg (~8 lbs) with semaglutide 1.0 mg and 2.7 kg (~6 lbs) with semaglutide 0.5 mg.
  • GI tolerability results were relatively good in this trial, but there were a few minor blips in the safety and adverse event profile. Rates of nausea (20%-24% of patients taking semaglutide) were comparable with those for other GLP-1 agonists, and better than we have seen from other semaglutide trials. The vast majority of nausea cases on both doses were categorized as mild, and especially for the lower dose, most resolved by week 15. Still, a not-insignificant proportion of patients started on semaglutide discontinued treatment due to adverse events (6.3% in the 0.5 mg group and 5.4 in the 1.0 mg group vs. 2.3% on placebo). Many patients will tolerate some amount of transient nausea if a drug’s glucose lowering efficacy – and especially its weight loss effects – are strong, but the higher rates of nausea seen elsewhere in the clinical development program for semaglutide make GI tolerability an item to watch as more phase 3 results comes out.
    • There was a slight imbalance in serious adverse events, with 5.4%-5.5% with semaglutide and 3.9% with placebo. Interestingly, there were four-to-zero imbalances in both gallstones and malignant neoplasms not in favor of semaglutide. The cancer cases were distributed across cancer types, and it is worth remembering that there were twice as many patients on semaglutide as on placebo in the trial. Gallstones are a risk factor for acute pancreatitis, but no cases of pancreatitis were reported in the trial. There was, however, the usual increase in pancreatic enzymes seen with most GLP-1 agonists. We wouldn’t ring any alarm bells yet, but we will be looking closely at the malignancy and gallstone numbers for other SUSTAIN trials.
  • What do the positive topline results from LEADER, the cardiovascular outcomes trial for semaglutide’s predecessor Victoza (liraglutide), mean for semaglutide? In terms of efficacy and weight loss, semaglutide appears to be delivering on its next-gen potential, with the caveat that there have not been head-to-head trials vs. Victoza. However, in Victoza’s favor, it is hard to beat demonstrated cardioprotection as a selling point. Interestingly, while Novo Nordisk ran LEADER as a fairly long and robust CVOT with the potential to demonstrate cardioprotection, SUSTAIN 6 (the CVOT for semaglutide) is remarkably short, with a registered start-to-finish more than two years shorter than LEADER’s – SUSTAIN 6 was actually completed earlier this year, and results may not follow too far behind those from LEADER. Since SUSTAIN 6 may not be long enough to demonstrate cardioprotection, Victoza may have a perpetual upper hand on the cardiovascular outcomes portion of its label. Such a scenario (especially if oral semaglutide becomes available) could slightly diminish injectable semaglutide’s relative appeal to payers, patients, and providers, though we still believe that those who take a look at the data will see semaglutide as the most effective agent out of Novo Nordisk’s GLP-1 agonist portfolio.

Questions and Answers

Q: So do you think there will be no cardiovascular outcomes benefit with this agent, unlike what we have with liraglutide?

A: Time will tell.

Q: Can you comment on what looked to be a minimal dose response between the two semaglutide doses?

A: It is probably multifactorial, and this is obviously speculative, but it may just be this specific population of patients. With the efficacy of this drug, and the fact that it turns itself off when glucose levels are normal, we may be getting patients down to an A1c level where there is not a lot of buffer to go lower. The SUSTAIN program is comprehensive, with different populations being studied, and future trials may show more of a dose response.

Q: How does the decrease in A1c compare with that for other GLP-1 agonists?

A: That would be comparing apples and oranges. This was a robust reduction in this patient population, but we don't yet have data on any direct comparisons.

Q: Could you discuss the four cases of malignancy seen in the semaglutide arm?

A: Yes, there were two in each semaglutide group, with no consistency across organ systems.

Q: Could you discuss injection site reactions and blood pressure data?

A: There were no statistically significant differences in either systolic or diastolic blood pressure. Injection site reactions were extremely rare, but I do not recall the specific numbers.

Q: Isn’t the lack of reduction in blood pressure different than what we see with other agents in the class?

A: There was a trend towards lower blood pressure, but it was not statistically significant in this study.


-- by Melissa An, Adam Brown, Emily Regier, Ava Runge, Manu Venkat, and Kelly Close