American Diabetes Association 77th Scientific Sessions

June 10-13, 2017; San Diego, CA; Day #3 Highlights – Draft

Executive Highlights

Greetings from sunny San Diego where the whirlwind of ADA 2017 is in full swing! This report details our top 17 highlights from day #3 of the meeting. Our therapy coverage is headlined by a batch of new LEADER data and suggestive but not conclusive results from the REMOVAL trial investigating metformin’s glycemic and CV effects in type 1 diabetes (the largest and longest cardiovascular study ever conducted in type 1 diabetes). On the tech front, Drs. Bill Polonsky and Jeremy Pettus debated the proposition of CGM as a standard of care in type 2 diabetes, and we got our second look at impressive Medtronic 670G consumer training phase data. Check out our highlights reports from day #1 and day #2 for a full picture of our learnings thus far, and take a look at what the next few days of ADA holds with our ADA 2017 resource hub, complete with category documents, our top picks for oral presentations, and a short list for the posters we’re most excited about.

Diabetes Therapy Highlights

1. In a highly-anticipated and absolutely packed session, Drs. Richard Pratley and Stephen Bain shared new cardiovascular, microvascular, and glycemic analyses from the LEADER cardiovascular outcomes trial (CVOT) for Novo Nordisk’s GLP-1 agonist Victoza (liraglutide). The analyses underscored the robust nature of liraglutide’s benefit even after adjusting for several factors. We also saw impressive A1c target achievement, delay to treatment intensification, and patient-reported quality of life data.

2. Full results from the REMOVAL trial investigating metformin’s glycemic and CV effects in type 1 diabetes were presented at an afternoon symposium, to an auditorium much less crowded than we expected considering the impact of cardiovascular disease in type 1 diabetes

3. In back-to-back presentations, Drs. Lawrence Leiter and Robert Henry shared findings on the lipid-lowering efficacy of Sanofi/Regeneron’s PCSK9 inhibitor Praluent (alirocumab) in people with diabetes specifically. Dr. Leiter presented results from type 2 diabetes participants in ODYSSEY DM-INSULIN trial of those with diabetes on insulin, which found a comparable LDL cholesterol lowering efficacy for alirocumab to previous trials and no new safety issues. Dr. Henry presented results from ODYSSEY DM-DYSLIPIDEMIA, which notably utilized a primary endpoint of non-HDL cholesterol. The ODYSSEY DM program is the first ever to investigate a PCSK9 inhibitor agent specifically in a diabetes patient population, so you can see why we’re very excited by theseis data.

4. AZ presented new one-year DURATION-8 results, demonstrating that the combination of AZ’s SGLT-2 inhibitor Farxiga (dapagliflozin) and once-weekly GLP-1 agonist Bydureon (exenatide) maintained improvements in glycemic control, body weight, and SBP compared to either drug alone.

5. Lexicon presented data on its SGLT-1/2 inhibitor sotagliflozin in type 1 diabetes in two late-breaking posters. One detailed the inTandem2 study demonstrating significant A1c reductions of 0.36% (200 mg dose) and 0.35% (400 mg dose) with sotagliflozin vs. placebo (p<0.001) in patients with type 1 diabetes on optimized insulin regimens (baseline A1c = a very low 7.7-7.8% after insulin optimization). The second detailed the JDRF-partnered study of sotagliflozin in young adults with poorly controlled type 1 diabetes; it failed to demonstrate significant A1c reductions vs. placebo but showed suggestions of improvement in other clinically relevant endpoints. From our view, this was the perfect trial to show why outcomes beyond A1c are critical.

6. Dr. Ildiko Lingvay (University of Texas Southwestern Medical Center, Dallas, TX) presented a meta-analysis of the SUSTAIN 1-5 clinical trials, demonstrating that Novo Nordisk’s once-weekly GLP-1 agonist semaglutide provided clinically meaningful and superior reductions in body weight across all five studies, versus a range of tested comparator drugs and in a broad range of patient populations.

7. Sanofi presented data from the DELIVER 3 study demonstrating that switching to Toujeo (insulin glargine U300) vs. other basal insulins was associated with significantly less hypoglycemia and similar glycemic control among older patients (≥65 years) with type 2 diabetes.

Diabetes Technology Highlights

1. In an instructive and entertaining session, Drs. Bill Polonsky and Jeremy Pettus debated use of CGM in type 2 – MDI, basal-only, and non-insulin – concluding it “will eventually become the standard of care for type 2 diabetes, especially as the technology becomes easier to use and less costly.” Still, the two debated persuasively in a pro/con format – see all the arguments below.

2. Yale’s Dr. Eda Cengiz shared glycemic outcomes from the MiniMed 670G Customer Training Phase, spanning March-May 2017 in 730 people (N=24,000+ patient days!). She emphasized strong real-world alignment with the pivotal data, including 74% time-in-range in Auto Mode (vs. 72% in the pivotal) and only 2% <70 mg/dl (vs. 3% in the pivotal). Mean glucose was 151 mg/dl (vs. 150 mg/dl in the pivotal), marking a 9-mg/dl improvement over manual mode.

3. A late-breaking poster detailed the continued disruption of the CMS Competitive Bidding Program – notably, the acclaimed set of authors estimate that, as of January 2014, over 90,000 insulin-treated people were calculating their insulin doses with partial/no SMBG. This is plainly unacceptable – what will it take to repeal a policy that is so clearly detrimental?

4. In a most valuable talk on clinical decision support tools, IDC’s Dr. Rich Bergenstal shared positive views on CGM standardization and the MiniMed 670G, highlighted the balance between population health and personal care (fascinating!), and commented on the amount of CGM data needed to make a clinical decision (two weeks is sufficient).

5. Harvard’s Dr. Frank Doyle peered into his crystal ball and predicted that fully closed loop (no meal announcement), embedded algorithms, adaptive control, faster insulin, and multiple sensors will all be a reality in the next five years.

6. Yale’s Dr. Stu Weinzimer reviewed some of the latest literature in closed loop systems beyond current hybrid models (including DIY) and introduced Barbara Davis Center’s Ms. Laurel Messer’s “CARE” plan for the application of automated insulin delivery to clinical care.

Other Highlights

1. In this year’s inspiring Banting Medal for Scientific Achievement lecture, Columbia’s Dr. Domenico Accili presented his vision of a revamped toolkit for diabetes care in 2021, 100 years after the discovery of insulin: Prevention of beta cell de-differentiation, a gut-targeted Foxo1 inhibitor to coax gut cells into producing insulin in a glucose-dependent manner, and selective insulin sensitizers.

2. A dQ&A, Close Concerns, and diaTribe Foundation late-breaker revealed the importance of time-in-range in patients’ daily lives – and how current therapies are not delivering on this metric.

3. Dr. Ann Albright, director of the CDC’s Division of Diabetes Translation, provided insight on her approach to sharing information about diabetes risk prevention. Acknowledging that robust research on this subject is scarce, Dr. Albright emphasized that one thing is clear: “we need to address prevention to a much greater degree than we currently are.”

4. Kicking off a lively morning symposium on the National Diabetes Education Program, Dr. Linda Siminerio (University of Pittsburgh Diabetes Institute, Pittsburgh, PA) underscored the importance of transitioning from traditional, paternalistic provider-driven decision-making to a more collaborative shared decision-making process.

Table of Contents 

Diabetes Therapy Highlights

1. New Cardiovascular, Microvascular, and Glycemic Analyses from LEADER

In a highly-anticipated and absolutely packed session, Drs. Richard Pratley and Stephen Bain shared new cardiovascular, microvascular, and glycemic analyses from the LEADER cardiovascular outcomes trial (CVOT) for Novo Nordisk’s GLP-1 agonist Victoza (liraglutide). On the CV side, Dr. Pratley showed that the robust CV benefit was retained even inclusion of recurrent CV events and after adjusting for concomitant therapy, severe hypoglycemia, and A1c. On the renal side, Dr. Bain similarly underscored the robust nature of the renal benefit demonstrated in the trial. Additionally, he shared impressive and exciting data demonstrating liraglutide superiority in A1c target achievement, delay to treatment intensification, and patient-reported quality of life. See our detailed discussion and commentary below for full results from this trial that just keeps on giving valuable data back.

2. Suggestive but Not Conclusive Findings from JDRF REMOVAL Study, Longest/Largest Trial of Metformin in Type 1 Diabetes

Full results from the JDRF-sponsored REMOVAL trial investigating metformin’s glycemic and CV effects in type 1 diabetes were presented at an afternoon symposium, to an auditorium much less crowded than we expected considering the impact of cardiovascular disease on life expectancy in type 1 diabetes.. Adults ≥40 years-old with type 1 diabetes (n=428) and at least three CV risk factors were randomized to twice-daily metformin at a 1,000 mg dose (n=219) or to placebo (n=209), both on top of standard of care. Treatment and data collection continued for three years, following an initial three-month run-in period of insulin optimization. From a baseline of 8%, patients on metformin experienced a mean A1c decline 0.24% greater than patients on placebo after three months (p<0.0001). After 36 months, this treatment difference was 0.13% in favor of metformin (p=0.006), though A1c had risen above baseline for both groups. While statistically significant, the authors of the paper (published online in the Lancet Diabetes & Endocrinology just as the symposium began) conclude that REMOVAL results do not support the assertion in current treatment guidelines that metformin causes a clinically-meaningful improvement in glycemic control for people with type 1 diabetes. Moreover, metformin did not meet the study’s primary endpoint of rate of progression of mean far wall cIMT (carotid intima media thickness), a proxy for CV outcomes. Over the course of the three-year study, mean cIMT increased at a rate of 0.006 mm/year in the metformin-treated group vs. a faster 0.010 mm/year rate of increase in the placebo group, though this trend did not reach statistical significance (p=0.1664). The related tertiary outcome of maximal cIMT , included as an outcome in REMOVAL due to its early mirroring in of later cardiovascular benefit in DCCT-EDIC, increased at a significantly slower pace in the metformin-treated group (0.012 mm/year) than the placebo group (0.25 mm/year; p=0.0093). Dr. Naveed Sattar delivered perspective remarks during the symposium, articulating that REMOVAL data is “suggestive but not conclusive” regarding metformin’s potential CV benefits in type 1 diabetes. During the subsequent panel discussion, other thought leaders including Dr. Julio Rosenstock expressed a more distinctly negative view of the data. We’re disappointed that metformin didn’t show more conclusive efficacy in REMOVAL, especially since it is a widely-accessible generic drug that many patients with type 1 diabetes consider as an adjunct to insulin, and there are so few data even on proxies of  cardiovascular disease in type 1 diabetes. We unpack the data in great detail below, and also share many more thoughts on what this all means  (should we be investing in metformin for type 1, when we see potential in phase 3 programs for SGLT-2 inhibitors in type 1?) – scroll down to our detailed discussion and commentary section.

One last thought: We found the press release announcing REMOVAL publication to be misleading, since the title suggests a direct link between metformin and decreased risk for heart disease, but REMOVAL was not a cardiovascular outcomes trialand the main content only mentions secondary/tertiary endpoints. We eagerly await far more commentary from thought leaders in the field as the dust settles.

3. First Dedicated Studies of a PCSK9 Inhibitor in Diabetes Report Compelling Positive Data

In back-to-back presentations, Drs. Lawrence Leiter and Robert Henry shared findings on the lipid-lowering efficacy of Sanofi/Regeneron’s PCSK9 inhibitor Praluent (alirocumab) in people with diabetes specifically. Dr. Leiter first presented the ODYSSEY DM-INSULIN trial (n=517), which demonstrated a 49% decrease in LDL cholesterol with alirocumab in patients with type 2 diabetes on insulin (analysis included only the 441 patients with type 2 diabetes – the trial also included 76 patients with type 1 diabetes and we’re working to find out if the effect was the same).  Overall, this LDL cholesterol reduction is consistent with what we’ve seen in previous ODYSSEY trials I in the general ASCVD population (with and without diabetes) – this is reassuring to support efficacy in people with diabetes, though we of course would’ve loved to see an even greater benefit in those with diabetes given the high residual CV risk. Next, Dr. Henry took the stage to present results from ODYSSEY DM-DYSLIPIDEMIA, which compared alirocumab twice-weekly (n=276) vs. usual care (n=137) in adults with type 2 diabetes and mixed dyslipidemia (defined by non-HDL ≥100 mg/dl, triglycerides between 150-500 mg/dl, and atherosclerotic CV disease or at least one other CV risk factor). After 24 weeks of treatment, participants in the alirocumab arm experienced a mean 37% reduction in non-HDL cholesterol vs. a 5% reduction in the placebo arm (p<0.0001), translating into a treatment difference of 33%. Baseline non-HDL was 155 mg/dl in the Praluent group, 162 mg/dl in the usual care group. Notably, this was the first major clinical trial to use change in non-HDL as its primary endpoint – during his talk on study design earlier in the symposium, Dr. Dirk Müller-Wieland contextualized this decision around an increasing number of professional organizations (the National Lipid Association, the European Society for Cardiology, and others) recommending non-HDL as a target for best practice lipid management. For more, including additional endpoint and safety data, see our detailed discussion and commentary section below.

4. Combination Therapy of AZ’s Farxiga and Bydureon Maintains Improvements in Glycemic Control and Body Weight Compared to Either Monotherapy in DURATION-8 Extension

AZ presented new one-year DURATION-8 results, demonstrating that the combination of AZ’s SGLT-2 inhibitor Farxiga (dapagliflozin) and once-weekly GLP-1 agonist Bydureon (exenatide) maintained improvements in glycemic control, body weight, and SBP compared to either drug alone. These results represent the 24-week extension from the promising 28-week DURATION-8 results presented at this past EASD meeting. In this double-blind, multicenter study, adults with type 2 diabetes were randomized to either exenatide plus dapagliflozin, exenatide alone, or dapagliflozin alone. Of the initial 695 patients randomized, 564 (81%) completed the full 52-week treatment period. The findings demonstrated that at week 52, the combination therapy group continued to achieve greater reductions in A1c, FPG, 2-hour PPG, body weight, and SBP compared to either treatment group alone. Specifically, the -1.8% A1c reduction for exenatide/dapagliflozin dual therapy significantly exceeded those of exenatide (-1.4%; p<0.01) and dapagliflozin (-1.2%; p<0.01) alone. Notably, Body weight reductions were -3.3 kg for the dual therapy arm, significantly greater than -1.5 kg for exenatide monotherapy (p<0.001) and -2.3 kg for dapagliflozin monotherapy (p<0.001). SBP reductions were 4.5 mmHg, 0.7 mmHg (p<0.001), and 2.7 mmHg, respectively. Regarding safety and tolerability, the combination therapy was well tolerated, with slightly more adverse events vs. either drug alone (66.2% vs. 62.2% [exenatide] and 61.8% [dapagliflozin]), though the three groups had comparable rates of serious adverse events. In general, patients treated with exenatide unsurprisingly reported more GI adverse events. These results confirm the promising findings that made waves at last EASD and the excitement surrounding the GLP-1/SGLT-2 combination approach. As both of these classes have the potential for CV risk reduction, we would be interested to see if a combination therapy could provide further additive benefits. For more on DURATION-8 and AZ’s latest, please see our EASD 2016 and 1Q17 coverage.

5. Promising Efficacy Data for Lexicon’s SGLT-1/2 Inhibitor Sotagliflozin in Type 1 Diabetes

Lexicon presented data on its SGLT-1/2 inhibitor sotagliflozin in type 1 diabetes in two late-breaking posters: one on the inTandem2 study in the general type 1 diabetes population and another (partnered with JDRF) in young adults with poor glycemic control.

  • The inTandem2 study demonstrated significant A1c reductions of 0.36% (200 mg dose) and 0.35% (400 mg dose) with sotagliflozin vs. placebo (p<0.001) in patients with type 1 diabetes on optimized insulin regimens (baseline A1c = 7.7-7.8% after insulin optimization). The company reported topline results from the study in December. The double-blind trial randomized 782 patients with type 2 diabetes to receive either placebo, 200 mg sotagliflozin, or 400 mg sotagliflozin for 24 weeks in addition to optimized insulin therapy. The study also included a double-blind long-term extension period of 24 weeks that is ongoing. In addition to the primary endpoint of A1c reduction at 24 weeks, investigators assessed a secondary endpoint of “net benefit”: the percentage of patients with A1c <7% at 24 weeks and no episodes of severe hypoglycemia or DKA during the study period. This endpoint was achieved by 32% of patients in both sotagliflozin groups vs. 15% of patients in the placebo group (p<0.001). While we appreciate seeing composite endpoints like these included in trials, as they are often more clinically relevant than A1c reductions alone, we understand from a regulatory perspective, there may need to be work standardizing them. In this case, while the significant difference vs. placebo is encouraging, the fact that less than a third of treated patients achieved the net benefit goal is somewhat disappointing. As for specific safety endpoints, rates of severe hypoglycemia were 3.8% with 200 mg sotagliflozin, 2.3% with 400 mg sotagliflozin, and 2.7% with placebo. DKA rates were 0.4% with 200 mg sotagliflozin, 1.1% with 400 mg sotagliflozin, and 0% with placebo.
  • The JDRF-partnered study of sotagliflozin in young adults with poorly controlled type 1 diabetes failed to demonstrate significant A1c reductions vs. placebo but showed promising improvements in other clinically relevant endpoints. Lexicon announced topline results from the study in December. The trial enrolled 87 patients age 18-30 with type 1 diabetes and an A1c ≥9%. Patients were randomized to receive either 400 mg sotagliflozin or placebo for 12 weeks in addition to insulin; the primary endpoint was change in A1c at 12 weeks. Sotagliflozin produced placebo-adjusted A1c reductions of 0.35% (baseline = 9.7-9.9%) and the difference between groups was not statistically significant (p=0.10). Because the primary endpoint was not significant, p-values for secondary endpoints could not be used to declare statistical significance. However, sotagliflozin did appear to produce improvements vs. placebo on postprandial glucose, body weight, time in range, and A1c in patients with a baseline A1c of 9-10%. In addition, approximately 8 times as many patients in the sotagliflozin group achieved the “net benefit” endpoint of A1c <7% at 12 weeks with no severe hypoglycemia or DKA compared to the placebo group, though the absolute percentage of responders was low in both groups (16% vs. 2%). We were especially impressed by the CGM data from the study – sotagliflozin therapy produced a one-third increase in time spent in range of 70 mg/dl-180 mg/dl. This study is a good example of the limitations of using A1c as the sole metric of efficacy for diabetes drugs, as other parameters like time in range and weight loss are likely more relevant for many patients. While there is almost exclusive focus on A1c at the regulatory level (though JDRF and The diaTribe Foundation among others are actively advocating for regulatory consideration of outcomes beyond A1c), it’s important to note that this was a phase 2, non-pivotal trial in an extremely challenging population. See our discussion of the topline results for more of our thoughts on this study.

6. Semaglutide Provides Superior Body Weight Reduction across All SUSTAIN 1-5 Clinical Trials

Dr. Ildiko Lingvay (University of Texas Southwestern Medical Center, Dallas, TX) presented a meta-analysis of the phase 3 SUSTAIN 1-5 clinical trials, demonstrating that Novo Nordisk’s once-weekly GLP-1 agonist semaglutide provided clinically meaningful and superior reductions in body weight across all five studies, versus a range of tested comparator drugs and in a broad range of patient populations. As a reminder, SUSTAIN 1 tested semaglutide against placebo (n=388; 30 weeks) in treatment-naïve patients, SUSTAIN 2-4 tested semaglutide head-to-head against other diabetes drugs in patients on metformin (sitagliptin in SUSTAIN 2 [n=1,231; 56 weeks], exenatide once-weekly in SUSTAIN 3 [n=813; 56 weeks], and insulin glargine in SUSTAIN 4 [n=1,089; 30 weeks]), and SUSTAIN 5 tested semaglutide against placebo as an add-on therapy in patients on metformin and insulin (n=397; 30 weeks). Average A1c and BMI at baseline were similar across all five studies, ranging from 8.1-8.4% and 32-34 kg/m2. Impressively, Dr. Lingvay outlined that significantly greater absolute reductions in body weight were seen with semaglutide across the entire SUSTAIN program. Participants on 1.0 mg semaglutide lost an average of 4.5 kg (9.9 lbs) in SUSTAIN 1 (versus -1 kg [2.2 lbs] with placebo), -6.1 kg (13 lbs) in SUSTAIN 2 (versus -1.9 kg [4.2 lbs] with sitagliptin), -5.6 kg (12 lbs) in SUSTAIN 3 (versus -1.9 kg [4.2 lbs] with exenatide once-weekly), -5.2 kg (11.5 lbs) in SUSTAIN 4 (versus +1.2 kg [2.6 lbs] with insulin glargine), and -6.4 kg [14 lbs] in SUSTAIN 5 (versus -1.4 kg [3.1 lbs] with placebo). Across the entire SUSTAIN program, participants on 1.0 mg semaglutide also achieved >10% weight loss significantly more often than those in the comparator group. The proportion of patients achieving this weight loss goal reached 13% in SUSTAIN 1 (versus 2% with placebo), 24% in SUSTAIN 2 (versus 3% with sitagliptin), 21% in SUSTAIN 3 (versus 4% with exenatide once-weekly), 16% in SUSTAIN 4 (versus 2% with insulin glargine), and 26% in SUSTAIN 5 (versus 3% with placebo). Dr. Lingvay alluded that longer-term data from the SUSTAIN 6 trial demonstrates that semaglutide’s impressive weight loss effects can be sustained for at least 2 years (more on this coming soon in our coverage of poster 1125-P) and that these weight loss effects are accompanied by significant reductions in A1c as well (results are detailed in poster 1124-P) Given that ~90% of people with type 2 diabetes struggle with overweight or obesity, weight management is a crucial advantage in a potential new diabetes medication, and all this is in addition to semaglutide’s added ability to reduce the risk of cardiovascular death, as demonstrated in the SUSTAIN 6 CVOT. Given this extremely promising clinical profile, this candidate, we expect the product is poised to perform very well in the booming GLP-1 agonist market. An FDA decision and an EU CHMP opinion for the submissions of once-weekly, injectable semaglutide for type 2 diabetes are expected in 4Q17 according to Novo Nordisk’s last earnings update. A more patient-friendly oral version of semaglutide remains in phase 3 development. 

7. DELIVER 3 study demonstrates significantly less hypoglycemia with no loss of glycemic control among older patients with type 2 diabetes who switched their basal insulin to Toujeo vs. other insulins

Sanofi presented data from the DELIVER 3 study demonstrating that switching to Toujeo (insulin glargine U300) vs. other basal insulins was associated with significantly less hypoglycemia and similar glycemic control among older patients (≥65 years) with type 2 diabetes. DELIVER 3 was a retrospective cohort analysis of de-identified patient-level EMR data in the US. The analysis included patients who switched to either Toujeo (n=468) or another basal insulin (n=1,142) between March 1, 2015 and March 31, 2016. Mean age was 71.8 years in the Toujeo group and 73.1 years in the comparator group; baseline A1c was 8.5% and 8.3% in the respective groups (both statistically significant differences). After six months, both cohorts achieved similar A1c reductions and a similar percentage of patients with A1c <7% and <8%. However, patients in the Toujeo cohort were 50% less likely to experience hypoglycemia during the follow-up period compared to the comparator cohort after adjustment for confounders (p=0.0002). The authors highlighted the fact that these results reflect Toujeo’s performance in real clinical practice, without the constraints of a randomized controlled trial. They also identified a number of limitations: the potential for selection bias, underreporting of hypoglycemia, coding errors, lack of adherence, and a lack of data on insulin dose, reasons for switching, duration of disease, and discontinuation rates. With these substantial limitations as caveats (the lack of dosing information seems particularly problematic), these results should help Sanofi build its case that Toujeo offers clinically relevant benefits over other basal insulins.

Diabetes Technology Highlights

1. Drs. Polonsky and Pettus Debate CGM in Type 2 Diabetes (MDI, Basal-only, Non-Insulin), Concluding It Will Eventually Become the Standard of Care

In what was a both instructive and entertaining session of ADA 2017, Drs. Bill Polonsky (BDI) and Jeremy Pettus (UCSD) debated use of CGM in type 2 – MDI, basal-only, and non-insulin – concluding it “will eventually become the standard of care for type 2 diabetes, especially as the technology becomes easier to use and less costly.” The final slides shared six major points of agreement following a vociferous debate: (i) with proper support, CGM could become a powerful motivational tool; (ii) innovative training materials are needed; (iii) new methods for providing CGM feedback are needed; (iv) episodic use of CGM may be best for many; (v) much more evidence on CGM in type 2 is needed; and (vi) we need to determine which patient types will benefit (e.g., the disengaged, hypoglycemia prone, chronically poor glycemic control, selecting the best medication, etc.). Still, don’t let these concluding remarks fool you – this session included hearty and hilarious debate on both sides, with Dr. Pettus taking the “pro” CGM side for type 2s on MDI and basal-only, and then switching to the “con” side for type 2s not on insulin. (For the switchover, the two actually exchanged sides of the stage, and in a dramatic gesture, humorously swapped sportcoats.) We include their key arguments for and against CGM in the different type 2 groups below. Both remain “very convinced that CGM is awesome” and holds a lot of promise in type 2, and their arguments provided a nice lens as to how it might help and where it definitely needs to improve.

Type 2s on intensive insulin therapy:

  • Dr. Pettus was staunchly in favor of CGM use in type 2 patients on intensive insulin therapy, while Dr. Polonsky argued that it may not always be a home run. Dr. Pettus structured his case around four tenets, largely informed by data from the DIaMonD type 2 cohort: (i) CGM is not more costly than other therapies. SGLT-2 inhibitor empagliflozin costs $400/month, GLP-1 agonist liraglutide costs $720/month, and CGM costs just $445/month – “people often overestimate the cost of CGM,” said Dr. Pettus, “but it is coming down.” In addition, there are no side effects to CGM use, unlike with adjunctive therapies, and patients usually want to be able to stop taking drugs, not the other way around. Plus, Dexcom’s G5 is now covered by Medicare for type 1s and type 2s (though the details of administering coverage are still pending), and many hope other payers will follow suit. (ii) Type 2s will wear CGM. In the DIaMonD type 2 cohort, 93% of patients were still using CGM ≥6 days per week in month six – presumably, they continued to wear it because they saw a benefit (though a clinical trial effect should not be totally discounted, especially because the study screening required consistent CGM use in the run-in). (iii) Some argue that type 2s aren’t tech savvy enough for CGM, but DIaMonD showed equivalent A1c drops regardless of education (bachelor’s degree or not), age, and numeracy. (iv) Dispelling rumors that CGM requires intensive education and time from providers, participants in the CGM group in DIaMonD had four visits in six months, and were only handed a trifold handout for education materials. “I don’t think the people are getting a benefit because they came into the office, but because they see data in real time and live a healthier lifestyle.”
  • Dr. Polonsky also channeled DIaMonD to suggest that the benefit of CGM in hypoglycemia in type 2s on intensive insulin therapy is nil, and wondered if CGM is doable in all cases of MDI-using type 2s. In DIaMonD, he said, there was no impact of CGM on severe hypoglycemia because there were zero episodes over six months in either group – to be fair, patients with recurrent severe hypoglycemia were excluded from the study, but Dr. Polonsky’s point is that the point of glycemic benefit should be proven by studying the appropriate population. We agree that more extensive hypoglycemia would be good to have, though the A1c drop and higher time in range seen in DIaMonD are encouraging. In terms of feasibility, Dr. Polonsky cast doubt on Dr. Pettus’ claims that CGM is doable in the real world: He argued that (i) in the real world, contact with a provider is much less frequent than seen in DIaMonD, and patients may require more support; and (ii) numeracy may be a greater concern than one might think (A&W’s third-pounder sold less than McDonald’s quarter-pounder “because four is bigger than three.”) In addition, he suggested that CGM use may be useful for some, but perhaps not all patients, but providing close support and education may be of great benefit.

Type 2s on basal-only

  • The second question of the debate – whether CGM should be prescribed for people with type 2 diabetes on basal insulin – swung unsurprisingly toward the “pro” side. Dr. Pettus spent his allotted time debunking what he felt were the biggest myths about CGM in this population: (i) that basal insulin is perfectly well titrated with SMBG; (ii) that patients won’t do anything with the results; and (iii) that hypoglycemia is uncommon. Indeed, he was quite masterful in weaving his way through a host of literature, poking holes in each myth one by one: (i) highlighting the prevalence of overnight hypoglycemia in this population and the inability of SMBG to capture this window; (ii) citing the actionable decision-making that has been documented in type 2s on basal insulin and the lifestyle changes that can persist for years (Yoo et al. 2008; Vigersky et al. 2012); and (iii) pointing out the massively underreported prevalence of hypoglycemia and severe hypoglycemia in this population. Altogether, Dr. Pettus kept his conclusion very simple, “I honestly think CGM should be standard of care in this population.” On the other hand, Dr. Polonsky took a more measured approach in suggesting that we do not yet have the evidence to indisputably call CGM the tool of choice in this population. Instead, he pointed out that studies of CGM in type 2 patients on basal insulin have been confounded by the vast (bordering on unrealistic) support provided to patients in clinical trials. Until we have data specifically looking at the effect of CGM independent of the extra support patients in clinical trials receive, Dr. Polonsky suggested that he’d put his final judgment on hold – we’re not sure how this would be parsed out, since any clinical trial will include extra support.

Type 2s not on insulin:

  • Dr. Polonsky, switching to the pro side, argued that some of the poor outcomes in type 2 diabetes might relate to “perceived treatment efficacy” – patients feeling that a therapy is actually working helps build momentum and a sense of progress. In diabetes, he said, we want to “help people to see that their actions are having a positive, tangible difference. Then, they get enthused. This is our opportunity with CGM! Humans respond to short-term, powerful, positive reinforcers.” Dr. Polonsky memorably noted that “feedback is the most underused motivational tool we have in diabetes,” and CGM provides an unparalleled level of feedback in real time. He reviewed the oft-cited Vigersky et. al paper (2012) testing CGM in non-insulin users, noting that episodic CGM use may be the way to go for those not on insulin (e.g., quarterly or at critical times like diagnosis, during diabetes education, when medication changes are needed, etc.).
  • Dr. Pettus, though a strong CGM supporter, took the con side quite persuasively, highlighting the grim realities of current medication adherence: nearly 1/3 of diabetes prescriptions are never filled, and of those that are filled, adherence rates at one year are <50% (Fischer et al., J Gen Intern Med 2010). He added that since 2005, 40 different treatment options have been approved for type 2 diabetes, with “approximately no change in A1c.” “Is CGM the 41st new therapy to move the needle,” he asked, particularly when it is more complicated to use than a once-daily pill? Dr. Pettus added that CGM’s training complexity might be a tall order for PCPs, who take care of ~94% of patients with type 2 diabetes on orals and don’t have time to fill out prior authorizations and train patients. He also underscored the realities of clinical inertia – after reaching an A1c of 8%, average time to add a medication was over one year (Brown et al., Diabetes Care 2004). “That’s one year to do one click (in the EMR) to prescribe a drug. A complex and time consuming therapy like CGM (harder to prescribe) will not help overcome patient and physician inertia.” Reimbursement also doesn’t exist for CGM in this group, meaning in a best-case scenario a prescription would be denied.

2. MiniMed 670G Glycemic Outcomes From Customer Training Phase (n=730) Strongly Resemble Pivotal Trial, Including 74% Time-in-Range and 2% in Hypoglycemia

In her review of hybrid closed loop systems, Yale’s Dr. Eda Cengiz shared glycemic outcomes from the MiniMed 670G Customer Training Phase, spanning March-May 2017 in 730 people (N=24,000+ patient days). She emphasized strong real-world alignment with the pivotal data, and we’ve enclosed the outcomes below using Medtronic’s own slide from yesterday’s ADA 2017 Analyst Briefing slide deck. (Dr. Cengiz’s numbers today were slightly different from what Medtronic shared yesterday, presumably due to rounding, mean vs. median, or different endpoints.) In the Customer Training Phase, 670G users spent 74% time-in-range in Auto Mode (vs. 72% in the pivotal), only 2% of the time <70 mg/dl (vs. 3% in the pivotal), 23% of the time >180 mg/dl (vs. 25% in the pivotal), and had a mean glucose of 151 mg/dl (vs. 150 mg/dl in the pivotal) – nice to see such alignment. Relative to manual mode, time-in-range on Auto Mode improved quite a bit more in the Customer Training Phase than in the pivotal – an 11-percentage point gain in real-world use (63%->74%) vs. a five-percentage-point gain in the pivotal (67%->72%). Mean glucose also improved by a nice 9 mg/dl in the Customer Training Phase (160->151 mg/dl) vs. no change in the pivotal. Customer Training Phase users spent a median 92% of the time in Auto Mode, also a rise from 87% in the pivotal. Guardian Sensor 3 wear remained strong at 95%, a positive sign for real-world use of Medtronic’s new sensor so far. Though these results are still in a very early adopter and enthusiastic population, they do inspire confidence as the device starts rolling out to 20,000+ priority access program participants (see coverage here, which only provided some of the data below).

  • Beyond the MiniMed 670G, Dr. Cengiz shared that she is working on a “personalized medicine” project testing a closed-loop module specifically for females with type 1 diabetes. We look forward to a future where a slew of algorithms are available to meet the needs of many different patients! On the horizon, Dr. Cengiz said the field still needs to improve on system training, more advanced technology with less burden, faster insulins, and more personalized management to reach a broad group of people with diabetes.

3. When Will the Madness End? Competitive Bidding Leaves 90k+ Dosing Insulin with Partial/No SMBG

This late-breaking poster shared new data documenting continued striking disruptions of Medicare beneficiary access to prescribed SMBG supplies following CMS’s expansion of its competitive bidding program (CBP) in 2013. The results follow up on the study’s original results (published in Diabetes Care) that showed an increase in mortality associated with the CBP in nine test markets in 2011. Scarily, it appears that the negative impact of the program has only worsened since 2013, when – as a reminder – CMS expanded the program nationally to both mail order and retail channels with lower reimbursement. This iteration of the study investigated changes in the acquisition of SMBG supplies by beneficiaries in those nine test markets (n=43,939) and all non-test markets (n=485,688) in the six months following the national CBP rollout and identified two major trends: (i) a significant increase in the percentage of beneficiaries who migrated from full SMBG to partial/no SMBG access in both test and non-test markets; and (ii) a significant increase in the percentage of insulin-treated beneficiaries with no record for SMBG (from 54.1% in January 2013 to 62.5% by December 2013, p<0.0001). Indeed, the authors estimate that as of January 2014, 37.5% (n=90,923) of insulin-treated beneficiaries were calculating their insulin dosage with partial/no SMBG. These results differ greatly from CMS’ April 2012 report on adverse outcomes associated with competitive bidding, which suggested that there was no disruption of access to supplies and no negative healthcare consequences associated with the program. However, the continued criticism and evidence to the contrary raises red flags for what is already a heavily scrutinized program. Our biggest question now that the evidence seems overwhelming … When will CMS actually listen? What will it take to reverse this policy?

4. Dr. Rich Bergenstal on CGM standardization, 670G, balance between population health and personal care

In a most valuable talk on clinical decision support tools, IDC’s Dr. Rich Bergenstal shared positive views on CGM standardization and the MiniMed 670G, highlighted the balance between population health and personal care (fascinating!), and commented on the amount of CGM data needed to make a clinical decision (two weeks is sufficient). We really enjoyed his views on the various needs clinical decision support tools must address – the balance between population health (A1c) vs. personal care (CGM); CGM “metrics” vs. CGM “profiles”; risk – “something needs to be done” vs. action – “here is what needs to be done”; and research & regulatory needs (publications, PIs, indications) vs. clinical management (therapy adjustments). Diving into CGM reporting, Dr. Bergenstal reminded attendees of Friday’s well-received consensus session, which honed in on the key CGM thresholds for reporting: <54, <70, 70-180, >180, and >250 mg/dl. Dr. Bergenstal noted that while the numbers are hopefully nailed down, the “terminology” is still open for debate – e.g., “Level 2 hypoglycemia” vs. “serious” vs. “clinically significant” vs. “take immediate action.” He noted that this will be discussed further at The diaTribe Foundation-organized meeting, “Glycemic Outcomes Beyond A1c: Standardization & Implementation,” in Bethesda, MD on July 21. Dr. Bergenstal also plugged an upcoming Diabetes Care publication from Dr. Roy Beck and colleagues, “The Fallacy of Average: How Using A1c Alone to Assess Glycemic Control Can Be Misleading” (see our take on this from last summer at diaTribe.org/BeyondA1c). Turning to AID, he said that ~70%-75% time-in-range (70-180 mg/dl) is a good goal for hybrid closed loop devices, given all the studies done to date. Dr. Bergenstal also showed new CareLink reports for the MiniMed 670G, which do a “nice job” of showing glucose profiles in Manual Mode vs. Auto Mode (similar to the pivotal trial plots, the profile shrinks in Auto Mode). Additionally, he covered the excellent one-page AGP report and noted the increasing commercial momentum, including in Abbott’s FreeStyle Libre, Dexcom’s Clarity (launched this week), Diasend/Glooko, and Roche. On Dr. Bergenstal’s recommended reading list? The Undoing Project by Michael Lewis, Reclaiming Conversation by Sherry Turkle, and the just-published Bright Spots & Landmines by our own Adam Brown.

  • “Two weeks of CGM data, most of the time, is representative of a longer period of time.” Dr. Bergenstal previewed poster 115-LB, which uses Senseonics 90-day pivotal trial data to answer an important question – how much CGM data is needed for a clinician? In comparing two weeks, one-month, and three months of CGM data, glucose patterns were very similar in all three cases. In our view, this is an especially good sign for clinicians using professional CGM.
  • The MiniMed 670G, Dr. Bergenstal joked, should be called “The Sleeping and Crying Machine. These parents are finally sleeping, and by and large, the fathers of these 13-17 year-olds come in – and while we’re talking about the numbers – they’re crying. ‘I finally have hope for my child.’”
  • Dr. Bergenstal characterized Dr. Bob’s Vigerksy’s Glucose Pentagon, combining five glycemic metrics into one graphic, “an amazing tool.” Poster 1049-P uses this glucose pentagon to summarize MiniMed 670G pivotal data – the pentagon area shrinks in comparing baseline to Auto Mode, indicating overall glycemic improvement. Dr. Bergenstal likes that the pentagon can compare pre-post treatment and also allows for comparison to people without diabetes. We like the cool factor of it, though it obviously doesn’t drive clinical decision making (how to change therapy) as much as the AGP.
  • www.agpreport.org is a nice website, showing all the AGP reports, including a new one for automated insulin delivery!

5. JDRF/NIH CL Research Meeting: Dr. Frank Doyle’s Predictions On the Next Five Years

Harvard’s Dr. Frank Doyle peered into his crystal ball and predicted that fully closed loop (no meal announcement), embedded algorithms, adaptive control, faster insulin, and multiple sensors will all be a reality in the next five years. None of the prophecies were earth-shattering by any stretch – we too believe they are within reach – but Dr. Doyle offered some interesting commentary on the near-term future of closed loop. He sees “glimmers of hope” in the algorithms arena, chiefly from the room for improvement in processing power embedded on a pump (“the [computer power] footprint on a pump pales in comparison to that on a cellphone”). Even so, he pointed out, existing capabilities can be extended by energy-aware algorithms or smart glucose monitoring. The algorithm will also have to take into account the PK/PD of faster insulins, if and when they do come to market (FIasp is already available outside the US and was resubmitted to FDA recently, with a decision expected in 3Q17). This shouldn’t be a problem ­– Dr. Doyle’s colleague Dr. Eyal Dassau has previously said that faster onset insulins would only make algorithms more effective. Lastly, he called for the integration of additional “sensors” into closed loop systems. The systems he is referring to, however, are not advanced biometric analyzers, but simple, everyday applications and consumer devices such as calendar events, health apps, fitness trackers, and GPS – these all produce rich and valuable information that should not be neglected for algorithm performance. Dr. Doyle reminded the audience that technology predictions are always a fraught area, and “the best way to predict the future is to create it.”

  • Our own Adam Brown brought up a critical question in Q&A: What features(s) will significantly drive closed-loop adoption to a broad population, and how should research dollars be allocated accordingly? For closed-loop technology to reach its full potential – improving lives at a massive public health scale – what features are needed? What must improve or change from current devices? Adam argued that cost will be the number one driver of this field’s expansion, given what we hear from patients (even in the US), studies of device barriers (e.g., Hood et al., Diabetes Care 2016), and the success of products like FreeStyle Libre. While burden and system performance are certainly critical, cost is the big driver of diabetes technology adoption in our view – the best closed-loop system is meaningless if patients/payers cannot afford it or see value in it. The closed-loop research community continues to focus a bit more narrowly on algorithm optimization, which is important for increasing the value of system, but is unlikely (in Adam’s view) to be the major driver of this field. We believe significantly less expensive closed loop systems would be a very high ROI area of research – and something fit for the community (NIH, JDRF, academia) to further address, as JDRF is doing in its T1D Outcomes Program, as we understand it. From our view, of course, showing the value is undoubtedly critical, since that lead to better reimbursement. We also like the point of getting all the technology right first. If cost is, by anyone’s definition, a critical metric for this field’s commercial viability, how do research priorities change? Costs might also influence study design: “what trials will most influence payers to cover closed loop” vs. “what trials would answer scientific/algorithm questions?” In a world where R&D/payer spending per patient has been shrinking, but where we’d like to see it increase, we see value in this area to improve public health on a broader scale. We think it would be valuable to look to the technology community, where over time, costs have dropped precipitously for various products – products, of course, that do not have costs like regulatory, admittedly. How could the field get to 80% of the benefits of closed loop systems with 20% of the costs? Would it ever be possible that the field make current pump+CGM therapy 10x less expensive – or, could value be much better demonstrated so that payers will be poised to deliver technology to a broader population?

6. Latest in Automated Insulin Delivery Beyond Hybrid Closed Loop; “CARE” Model to Apply Closed Loop In Clinical Care

Yale’s Dr. Stu Weinzimer reviewed some of the latest literature in closed loop systems beyond current hybrid models (from Harvard IP delivery, BU bihormonal, and OpenAPS) and introduced Barbara Davis’ Ms. Laurel Messer’s CARE plan for the application of automated insulin delivery to clinical care.

  • Like most physicians, Dr. Weinzimer sees the “Wild West of [DIY] automated insulin delivery” as a double-edged sword – on the one hand, it worries him that people are using unregulated technology to dose a potentially lethal drug, but on the other hand, the “positive disruption” is not lost on him. While there is without a doubt a risk to these systems, they have generated positive outcomes, largely anecdotally, and more recently in peer-reviewed literature: OpenAPS’ Ms. Dana Lewis et al. recently published a piece in JDST (a poster at ADA 2016) showing glycemic metrics for >100 OpenAPS users, comprising >250,000 total closed loop hours. For those of you keeping track, that’s at least 125,000 more hours than the Cambridge system has seen in patients. And the real-world data is encouraging: Relative to baseline, users of the system included in the study have seen a self-reported median 0.9% A1c drop (from a low baseline of 7.1%), and time in range (70-180 mg/dl) leapt from 58% to 81%. DIY systems do have risks, but so does dosing insulin every single day. Patients can iterate quickly and develop products that fit their needs outside of the traditional clinical trial and regulatory cycles.
  • BDC’s Ms. Laurel Messer has a publication in press (Pediatric Diabetes) explaining her “CARE” framework for the application of automated insulin delivery to clinical care. CARE is an acronym for Calculate (how does the system calculate insulin delivery?), Adjust (what are the adjustable components?), Revert (when should control be returned to the user?), and Educate (where does the user go for help?). Ms. Messer has over three years of experience with the Medtronic 670G hybrid closed loop system (read about some of her fantastic insights here). We’re eager to read more about the CARE model, but at a high level, we like that it emphasizes the human-machine interaction and the importance of graceful handoffs between closed loop and open loop. Dr. Weinzimer commented that it’s on industry to reach out to providers to facilitate smooth uptake for patients. 
  • Early in his talk, Dr. Weinzimer called attention to Dr. Eyal Dassau et al.’s recent Diabetes, Obesity, and Metabolism publication showing much improved glycemic control with IP vs. subcutaneous insulin delivery in the context of fully-automated artificial pancreas. Time in the tight 80–140 mg/dl range was 40% in the IP group vs. 26% in the subcutaneous group (p=0.03). Other outcomes, including mean glucose, time in a broader 70-180 mg/dl range, and time in hyperglycemia were all superior in the IP group – numbers aside, we were most struck by the blunting of postprandial spikes in the accompanying glucose profile graphs. Opinion leaders argue that IP delivery is more physiologic than subcutaneous delivery (most recently at a JDRF/HCT workshop), but what will it take to design a product that is affordable and desired by patients, particularly as improved subcutaneous options become more cost-effective, less invasive, and improve in patient/prescribing burden?
  • For insulin+glucagon bihormonal closed loop, Dr. Weinzimer displayed data from the latest BU/MGH publication showing home use of a bihormonal bionic pancreas decreases average blood glucose and time in hypoglycemia over the team’s 11-day multicenter study. Participants were randomly assigned to bionic pancreas (n=20) or usual care (n=19, conventional or sensor-augmented pump). Over the duration of the trial, the mean blood glucose was ~140 mg/dl in the bionic pancreas group, and ~162 mg/dl in the comparator arm. Time in hypoglycemia was also significantly lower in the bionic pancreas arm (0.6% vs. 1.9% time <60 mg/dl). The bionic pancreas has conferred very strong glycemic outcomes in studies (mean glucose + hypoglycemia), and the qualitative meal announcement (no carb counting) and easy initialization (based solely on patient weight) make it an appealing option for the user. Skeptics point out that the incremental value of glucagon vs. its cost is an unknown – though this is often the case with new innovations. As a reminder, an NIDDK-funded pivotal trial of the bihormonal bionic pancreas is slated for mid-2018, while the insulin-only version is expected to enter its pivotal trial in late 2017/early 2o18.

Other Highlights

1. Banting Medal for Scientific Achievement Lecture: Dr. Domenico Accili’s Hopes for Diabetes Care Toolkit in 2021

In this year’s inspiring Banting Medal for Scientific Achievement lecture, Columbia’s Dr. Domenico Accili presented his vision of a revamped toolkit for diabetes care in 2021, 100 years after the discovery of insulin: Prevention of beta cell de-differentiation, a gut-targeted Foxo1 inhibitor to coax gut cells into producing insulin in a glucose-dependent manner, and selective insulin sensitizers. These sound like ambitious goals, but Dr. Accili conveyed confidence-inspiring narratives around each. The meat of his talk focused on dispelling the perception that beta cell failure is a consequence of beta cell death. His data seem to suggest that beta cell dedifferentiation, not necessarily death, is actually to blame. In the islets of individuals with diabetes, beta cells are still alive, but they lose their expression of insulin and other beta cell-typical hormones and come to resemble progenitors – in fact, the number of hormone-negative cells in human islets can be as high as 30%, and the process of dedifferentiation is consistent with the clinical features of type 2 diabetes. Eventually, some of these even convert to glucagon-producing alpha-like cells. Dr. Accili sees a big opportunity here: If beta cells are not dead, but just quiescent as dedifferentiated of converted cells, then there’s a chance to restore beta cell health even after the onset of hyperglycemia. For type 1 diabetes, Dr. Accili’s lab has set sights on coaxing gut cells into producing insulin in a glucose-dependent fashion. Early studies found that knocking out FOX1 yields insulin positive cells in the gut, which exhibit glucose-dependent insulin secretion ex vivo. The strategy of converting intestinal cells into insulin-producing cells may be better suited to treat type 1 diabetes than stem cell-derived beta cells because the gut has “immune privilege” (the cells are less likely to be targeted as invaders), endocrine cells have a short half life in the gut, and so escape destruction, and gut cells continually regenerate, ensuring a renewable reservoir. In Dr. Accili’s opinion, “there’s no reason why this should not work.” Lastly, for type 2 diabetes, Dr. Accili explained that it is possible to modulate critical nodes of insulin signaling to dial up/down individual bio-responses, enabling selective reversal of insulin resistance – any therapeutic agent in this vein could reduce the burden on overwhelmed beta cells and preserve glycemia for extended periods of time. With a portfolio this lengthy and deep, we can’t think of anyone more deserving of this award than Dr. Accili – our only hope is that he sticks with it and delivers 2021: A Diabetes Odyssey.

2. Do Patients Feel Successful on Current Therapy? dQ&A Survey Says No. Does Time-in-Range Matter? Survey Says … Yes!

This study – authored by teams at The diaTribe Foundation, Close Concerns, and dQ&A – surveyed patients with type 1 and type 2 diabetes, emphasized the importance of time-in-range in patients’ daily lives – and how current therapies are not delivering on this metric. Results were collected by surveying members of the dQ&A Patient Panel, who were asked about the factors that have the greatest impact on their daily lives and potential drivers of improvement in mindset and diabetes management. The evaluation received responses from a remarkable 73% of those invited, totaling 3,461 people with type 1 and type 2 diabetes (n=1,026 and 2,435, respectively). Participants said that current therapies are coming up short in a number of areas, most notably in terms of helping patients achieve their desired time-in-range numbers, diet/exercise goals, and emotional well-being. Patients rated “time spent in the ideal blood glucose range” as having the biggest impact on their daily lives – even higher than the impact of A1c for type 1s (T1), type 2s on insulin (T2I), and type 2s not on insulin (T2NI) – but despite the importance placed on this metric, only 23% of T1, 25% of T2I, and 38% of T2NI reported that their current therapies are “very successful” at delivering in-range numbers (70-180 mg/dl). Similarly, a substantial proportion of patients (26% T1, 35% T2I, 50% T2NI) reported that a change in “diet and exercise” would have the biggest positive impact on their diabetes, likely resulting in a “big improvement” in their health. In spite of this, a minority of patients in each group (T1: 28%; T2I: 10%; and T2NI: 17%) said that their current therapies were “very successful” at “reaching, or keeping to, a healthy weight.” Finally, on psychological health, results showed that emotional well-being was disturbingly low across all three groups, with only a few patients rating therapies as “very successful” at achieving this outcome (22-34%). Taken together, these findings remind us that there is a significant need to more expansively evaluate what matters to patients and better align outcomes with patient priorities. We hope this work prompts future exploration of how such priorities can be incorporated into therapy development, regulatory decisions, and reimbursement. As a reminder, this data was presented in brief last year at the FDA Outcomes Beyond A1c Workshop. The diaTribe Foundation will propel the conversation forward at another FDA workshop on July 21st in Bethesda, MD. For those interested in a more detailed look at our data and conclusions or information on the upcoming gathering, please write us!

3. CDC’s Dr. Ann ALbright on How to Communicate About Diabetes Risk Perceptions and Prevention

The legendary Dr. Ann Albright, director of the CDC’s Division of Diabetes Translation, provided invaluable insight into her approach to sharing information about diabetes risk prevention. Though she acknowledged that robust research on this subject is scarce, Dr. Albright emphasized that one thing is clear: “we need to address prevention to a much greater degree than we currently are” – hear hear! There is clearly a large role for policy and community-level strategies in the promotion of diabetes prevention, but in the context of one-on-one conversations in the physician’s office, Dr. Albirght’s take-home message was the need for greater sensitivity to which conversation topics are most needed for people at different levels of diabetes risk – low (general information about healthy behaviors), medium (NDEP resources and recipe guides), and high risk (structured programs like the DPP, or medication like metformin).

4. The Importance of Shared Decision-Making

Kicking off a lively morning symposium on the National Diabetes Education Program, Dr. Linda Siminerio (University of Pittsburgh Diabetes Institute, Pittsburgh, PA) underscored the importance of transitioning from traditional, paternalistic provider-driven decision-making to a more collaborative shared decision-making process. Illustrating the need for a change in the way that patients and providers typically communicate, Dr. Siminerio reviewed the results of an Institute of Medicine survey quantifying Americans’ beliefs, attitudes, and preferences about healthcare. The findings were striking: 80% of people want their healthcare provider to listen to them, but only 60% say it happens; less than half of people report that their provider asks about their personal health goals; and 90% of people want their healthcare providers to act as a team, and yet only 40% say it happens. Follow-up studies further illustrate the negative consequences that can emerge from an environment that doesn’t encourage shared decision-making; patients without decision support are 60% more likely to change their mind about a treatment plan, 23% more likely to delay a healthcare decision, five times more likely to regret their healthcare decisions, and 19% more likely to blame their physician for bad outcomes. Against this backdrop of the clear need for greater shared decision-making in clinical practice, Dr. Siminerio reminded the audience that barriers to shared decision-making are not easily overcome. Healthcare providers may find it a challenge to their autonomy and often find it difficult to communicate nuanced data to people without a medical background, whereas patients often suffer from health literacy challenges and a lack of useful decision aides. Kudos to Dr. Siminerio for being such an advocated for better communication – congratulations to her and her NDEP colleagues for a 20th anniversary! We look so forward to returning in our full report to share many other valuable talks from this session.

Detailed Discussion and Commentary

Symposium: New Learnings from the Results of the Liraglutide Effect and Action in Diabetes – Evaluation of Cardiovascular Outcome (LEADER) Trial

New Learnings – Cardiovascular Outcomes

Richard Pratley, MD (Florida Hospital, Orlando, FL)

Dr. Richard Pratley shared new cardiovascular analyses from the LEADER cardiovascular outcomes trial (CVOT) for Novo Nordisk’s GLP-1 agonist Victoza (liraglutide). Initially presented to great fanfare at ADA 2016, LEADER (n=9,340) demonstrated a 13% risk reduction for the primary 3-point MACE endpoint (non-fatal MI, non-fatal stroke, and CV death) with liraglutide therapy compared to placebo (p=0.01 for superiority). Dr. Pratley delved into several additional analyses of the cardiovascular findings, underscoring the robust nature of the benefit. Notably, the primary endpoint included only time to first occurrence of a CV event – so a person who experienced two non-fatal MIs would only count as one “event” in the primary analysis. When recurrent CV events were included in the analysis (or, in other words, when looking at the total number of non-fatal MIs, non-fatal stroke, and CV deaths that occurred in the trial), the hazard ratio remained remarkably consistent in favor of liraglutide and highly statistically significant (HR=0.86, 95% CI: 0.78-0.95, p=0.004). As Dr. Pratley put it, this finding provides a better sense of the benefit of liraglutide on the overall burden of disease since, unfortunately, recurrent CV events do occur in far too many patients (and, indeed, those with a prior CV event are at higher risk for future events). All in all, this finding provides evidence of a CV benefit of liraglutide in even the highest risk populations.

  • Dr. Pratley examined the potential impact of concomitant medications on the primary outcome findings in LEADER. This is especially important as some have suggested that the LEADER findings may be driven by differences in CV medication or glucose-lowering therapy between the two arms. The investigators evaluated the impact of liraglutide on primary outcome event rate when adjusted for beta-blockers, ACE inhibitors, statins, or platelet aggregation inhibitors and the hazard ratios and 95% confidence intervals remained remarkably consistent despite these adjustments (see table below). Further, some have suggested that increased insulin, SU, or TZD usage in the placebo group during the trial could be driving the results by causing harm in the placebo arm. This hypothesis was not substantiated by the analyses presented by Dr. Pratley – the hazard ratio of benefit in favor of liraglutide also remained remarkably consistent regardless of concomitant antihyperglycemic therapy (ranging from 0.79 to 0.88, see table below for full results). This finding is particularly reassuring and further reinforces that the CV benefit of liraglutide is likely “real.” The LEADER team also examined the primary outcome results in patients who did and did not experience a severe hypoglycemia event and again found very consistent hazard ratios. We note that the confidence interval for the subgroup of those who did experience severe hypoglycemia was very, very large, likely a reflection of the relatively small number of patients who experienced this event, due in part to the impressive 31% hypoglycemia risk reduction associated with liraglutide.
  • Dr. Pratley noted that the CV benefit of liraglutide is not readily explained by differences in clinical and metabolic endpoints, despite the end-of-trial differences in A1c (0.4% favoring liraglutide, p<0.001), weight (2.4 kg [5.1 lbs] favoring liraglutide, p<0.001), systolic blood pressure (1.2 mmHg favoring liraglutide, p<0.001), and lipids. Looking to A1c specifically, the rate of primary endpoint occurrence was similar across subgroups when participants were stratified by in-trial A1c reduction from baseline at six months. The primary endpoint was experienced by 15% of liraglutide-treated patients with >1.5% A1c reduction (12% of placebo patients), 13% of patients with A1c reduction between 0.4% and 1.4% (12% of placebo patients), and 14% of patients with A1c reduction <0.4% (14% of placebo patients). The similarity in event rate across the three A1c reduction subgroups suggests that it’s unlikely A1c alone that’s driving the benefit, according to Dr. Pratley.
  • Overall, the additional analyses provided more information on what is not driving the CV benefit of LEADER than offering potential explanations. Given that the benefit does not appear to be driven by concomitant therapies or known metabolic characteristics of liraglutide, Dr. Pratley posited that the benefit is likely driven by “off-target” effects of GLP-1 agonism, whether acting directly on the heart or acting on the vasculature. That said, we’ve also seen some heterogeneity in the GLP-1 agonist class thus far – Novo Nordisk’s semaglutide also reported positive results in SUSTAIN 6, but Sanofi’s Lyxumia (lixisenatide), AZ’s Bydureon (exenatide once-weekly), and Intarcia’s ITCA 650 (implantable exenatide mini-pump) did not demonstrate CV superiority in their own CVOTs (though the last one was admittedly a much smaller trial designed for an easier safety threshold). Given that human GLP-1-based liraglutide and semaglutide differ substantially from the exendin-4-based GLP-1 agonists that have shared CV results thus far (and, some suggest Novo Nordisk’s compounds also differ substantially other human-based GLP-1 agonists in terms size, etc.), it’s quite possible that we’ll continue to see substantial heterogeneity in this class in our view.

Additional CV Analyses for LEADER

Analysis

Primary outcome hazard ratio (95% CI)

Unadjusted primary finding

0.87 (0.78-0.97)

Recurrent CV events included

0.86 (0.78-0.95)

Adjusted for beta-blockers at baseline

0.86 (0.77-0.96)

Adjusted for ACE inhibitors at baseline

0.87 (0.78-0.97)

Adjusted for statins

0.87 (0.78-0.97)

Adjusted for platelet aggregation inhibitors

0.87 (0.78-0.97)

Subgroup of those on insulin at baseline

0.88 (0.75-1.03)

Subgroup of those not on insulin at baseline

0.86 (0.74-1.01)

Subgroup of those never treated w/ insulin during trial

0.82 (0.68-0.98)

Subgroup of those never treated w/ SU during trial

0.79 (0.67-0.94)

Subgroup of those never treated w/ TZD during trial

0.87 (0.78-097)

Subgroup of those who did not experience severe hypoglycemia during trial

0.88 (0.78-0.98)

Subgroup of those who did experience severe hypoglycemia during trial

0.85 (0.52-1.39)

New Learnings – Secondary Outcomes

Stephen Bain, MD (Swansea University, UK)

Dr. Bain offered additional granularity on the impressive secondary microvascular benefit observed in the LEADER trial. At ADA 2016, the LEADER investigators shared that liraglutide therapy produced a 16% statistically significant improvement in time to first microvascular event (encompassing renal and ophthalmic adverse outcomes; 95% CI: 0.73-0.97, p=0.02), driven entirely by a 22% statistically significant improvement in renal outcomes (95% CI: 0.67-0.92, p=0.003). There was no statistically significant improvement in eye outcomes – the hazard ratio was 1.15 (95% CI: 0.87-1.52, p=0.33). Focusing on the renal outcome, Dr. Bain showed that the results remained virtually the same when adjusted for use of RAAS inhibitors at baseline (HR=0.78, 95% CI: 0.67-0.92, p=0.002). Expanding upon the urinary albumin-to-creatinine ratio (UACR) data presented at EASD 2016 (in which liraglutide produced a 19% reduction in urinary albumin-creatinine ratio, a measure of microalbuminuria [HR=0.81, 95% CI: 0.76-0.86]), Dr. Bain showed that the estimated treatment ratio was comparable regardless of whether patients’ had normal albuminuria (HR=0.86; 95%CI: 0.80-0.93), microalbuminuria (HR=0.76; 95% CI: 0.67-0.85), or macroalbuminuria (HR=0.87; 95% CI:0.72-1.06).

  • Dr. Bain also offered an exciting expanded look at the glycemic and patient-reported outcomes associated with liraglutide therapy in LEADER. Participants in the liraglutide arm of the trial were 79% more likely to achieve an A1c<8% and more than twice as likely to achieve an A1c <7.5% or <7% compared to placebo at three years. Even more impressively, liraglutide therapy was associated with a significant delay in time to insulin initiation, producing a 48% relative risk reduction in insulin initiation (HR=0.52, 95% CI: 0.48-0.58). The results were similar for time to initiation of insulin or of any new oral agent: liraglutide was associated with a 37% risk reduction for therapy intensification (HR=0.63; 95% CI: 0.59-0.68). We’ve long felt that earlier use of GLP-1 agonists could delay the need for treatment intensification (and the associated hypoglycemia, weight gain, etc. concerns associated with insulin) and we’re pleased to see long-term clinical evidence in support of this. We were also extremely pleased to see patient-reported quality of life results from one of the largest trials of a GLP-1 agonist to date. In LEADER, liraglutide demonstrated superiority in both the European Quality of Life (EQ-5D) index score (p=0.020) and the EQ-5D visual analog scale (VAS) score (p=0.019).

Symposium: Results of the JDRF Reducing with Metformin Vascular Adverse Lesions in Type 1 Diabetes (REMOVAL) International Multicenter Trial

Introduction, Study Rationale, and Design

Helen Colhoun, MD (University of Edinburg, UK)

The REMOVAL trial was published online in Lancet Diabetes & Endocrinology just as this symposium began, and Dr. Helen Colhoun opened by outlining motivations for the trial and reviewing study design. She highlighted a lingering and concerning problem in type 1 diabetes care – patients face a 2.5-4x risk for CV disease compared to a background population, without adequate adjunct therapies to address it. Metformin is prescribed for some type 1 patients and is recommended in ADA and other several treatment guidelines. However there have only ever been nine small studies, the largest of which studied was in 100 participants followed for one year (a total randomized follow-up in all studies of 192.8 patient years).  The evidence summarized in a meta-analysis published in 201o by Dr Colhoun and Dr Petrie It has demonstrated insulin-sparing effects, A1c-lowering efficacy, weight loss benefits, and possible A1c-lowering efficacy. There was and a slight impact on LDL in one previous study in which few patients were treated with statins. in a host of prior studies – each of these measures individually is correlated with CV outcomes  (DCCT/EDIC, for example, is the only study that has showeded a long-term benefit on macrovascular complications in type 1, stemming from intensive glucose-lowering, but risk factors other than glycemia are correlated with CV outcomes), so at the very least, metformin might lower frequency of CV events by impacting other risk bringing down the contributing factors or by other pleiotropic effects as in type 2 diabetes.  Ambiguity surrounding metformin’s mechanism of action means that much remains unknown about the molecule’s specific effects on different physiological systems. All of this possibility led the REMOVAL research team to select this well-known generic drug for a CV study. The group decided to look at an intermediate marker of CV risk, carotid intima media thickness (cIMT), as a proxy for potential long-term cardioprotection. Adults ≥40 years-old with type 1 diabetes (n=428) were randomized to twice-daily metformin at a 1,000 mg dose (n=219) or to placebo (n=209), both on top of standard of care (which included insulin titration/adjustments). Treatment and data collection continued for three years, following an initial three-month run-in period of insulin optimization. The primary outcome was rate of progression of mean far wall cIMT at baseline, 12 months, 24 months, and 36 months. Secondary outcomes included A1c, LDL cholesterol, albuminuria, weight loss, insulin dose, and endothelial function. Rate of progression to maximal cIMT was a pre-specified tertiary endpoint.

Study Population

Alicia Jenkins, MD (University of Sydney, Australia)

REMOVAL studied 428 middle-aged adults with long-standing type 1 diabetes at high risk of cardiovascular disease, as defined by three or more risk factors (including BMI >27 kg/m3, A1c >8%, known cardiovascular or peripheral vascular disease, current smoker, high blood pressure, high cholesterol or triglycerides, strong family history of cardiovascular disease, or diabetes duration >20 years). Baseline characteristics were well-balanced between the metformin and placebo-treated arms in terms of age (55 and 56 years), diabetes duration (33 and 34 years), BMI (28 kg/m3), A1c (8.1% and 8.0%), total daily insulin dose (0.63 and 0.68 units/day), frequency of prior cardiovascular disease (14% and 11%), blood pressure (systolic: 130 and 129 mmHg; diastolic: 73 and 72 mmHg), and cholesterol (LDL=85 mg/dl and HDL=62 mg/dl for both groups). Notably the study population had high rates of retinopathy: only 12% of participants in the metformin arm and 8% of participants in the placebo arm had no form baseline retinopathy, with the rest of the population showing high incidence of non-proliferative diabetic retinopathy (64% and 63%), inactive proliferative diabetic retinopathy (16% and 19%), and proliferative diabetic retinopathy (7% and 9%). On the renal front, the majority of participants in the metformin and placebo arms alike had normal eGFR (58% and 62%), though there was some incidence of stage 1 CKD (11% in both groups), stage 2 CKD (27% vs. 23%), and stage 3a CKD (3% and 4%). Furthermore, the study population had high usage of statins (82%), antihypertensives (73%), and anti-platelet drugs (39%).

Glycemia

Irene Hramiak, MD (St. Joseph’s Healthcare, Ontario, Canada)

Dr. Irene Hramiak took the stage to discuss the impact of metformin on A1c and insulin dose in the three-year REMOVAL trial. From a baseline of 8%, patients on metformin experienced a mean A1c decline 0.24% greater than patients on placebo after three months (p<0.0001). After 36 months, this treatment difference was 0.13% in favor of metformin (p=0.006), though A1c had risen above baseline for both groups. Dr. Hramiak characterized this effect as “small, but highly statistically significant,” calling attention to the steep slope of glucose-lowering in the first three months, even though the benefit appeared to dissipate with follow-up.

  • Average daily insulin dose grew for participants in the placebo arm but decreased for participants in the metformin arm over three years, culminating in a “modest” treatment difference of 1.9 units/day (p=0.0018). Of note, a majority of patients enrolled in REMOVAL (58%) were on a background of basal bolus therapy, while 33% were on an insulin pump, 3% were on twice-daily basal insulin, and 5% were on some other insulin regimen.
  • The hazard ratios for minor and severe hypoglycemia were 1.12 and 1.23 in favor of placebo at the 36-month mark, but neither of these reached statistical significance (p=0.259 and 0.442, respectively), leading Dr. Hramiak to conclude that metformin did not increase risk for hypoglycemia. While this is important data to report from a safety perspective, one of the reasons we’re pushing for better adjunct therapies in type 1 diabetes is that we want to lower a patient’s insulin dose and thereby reduce hypoglycemia – a point estimate in the “wrong direction” is somewhat disappointing in this context. We might’ve hoped for a significant hypoglycemia benefit in REMOVAL, but we recognize that insulin dose wasn’t drastically decreased with the addition of metformin by any means (as Dr. Julio Rosenstock argued during Q&A, a change of ~two units per day is “absolutely nothing”). All in all, these glycemia results were disappointingly neutral. We’re hesitant to give up on metformin’s role in type 1 diabetes care and treatment algorithms (see the Q&A section below for a deeper dive on this topic), but as acknowledged by the based on REMOVAL investigators, and contrary to current guidelines, metformin did not offer a clinically-meaningful benefit to A1c on top of insulin therapy.

Primary Endpoint

Nishi Chaturvedi, MD (University College London, UK)

REMOVAL’s primary endpoint was mean carotid intima media thickness (cIMT), a commonly-used surrogate marker for cardiovascular risk. cIMT essentially reflects the thickness of the inner two layers of the carotid arteries (atherosclerosis) and is easilycan be measured by careful and quality assured noninvasive ultrasound techniques. Over the course of the three-year study, mean cIMT increased at a rate of 0.006 mm/year in the metformin-treated group vs. a faster 0.010 mm/year rate of increase in the placebo group, though this trend did not reach statistical significance (p=0.1664). However, the related tertiary outcome of maximal cIMT increased at a significantly slower pace in the metformin-treated group (0.012 mm/year) than the placebo group (0.25 mm/year; p=0.0093).  In biological terms, mean cIMT reflects overall thickening of the wall of the artery, while maximal cIMT, favored by the DCCT-EDIC investigators, includes areas of plaque and focal thickening.

Clinical and Metabolic Outcomes

John Petrie, MD (University of Glasgow, Scotland)

Dr. John Petrie, primary author of the study, presented secondary endpoint data. From a baseline 84 kg (185 lbs), body weight decreased for patients on metformin and increased for patients on placebo, leading to a treatment difference of 1.2 kg (2.6 lbs) sustained after over three years (p<0.0001). Metformin also demonstrated positive effects on LDL, which fell 5 mg/dl in the treatment group but remained flat in the placebo group at year three (p=0.0117) and on eGFR – this measure of renal function rose sharply for people on metformin in the first three months, and after 36 months showed a treatment difference of 4 ml/min/1.73m2 in favor of metformin (p<0.0001). In less positive news, metformin was not associated with a statistically significant benefit to retinopathy or to endothelial function.

  • The signals for weight loss and lipid-lowering efficacy support metformin’s current niche in type 1 diabetes care, in that many HCPs prescribe the generic drug to their type 1 patients to mitigate insulin-induced weight gain in those with overweight or obesity. During Q&A, Dr. Partha Karr shared that these findings provide some reassurance on how he runs his practice and when he turns to metformin for type 1, which is most often for a patient who also wants to lose weight or improve cholesterol. Looking at REMOVAL through this lens, the results still spotlight metformin’s value and utility in type 1 diabetes management, although they don’t bring to light any new benefits to CV biomarkers or glycemia.

Safety

Martijn Brouwers, MD (Maastricht University Medical Center, Maastricht, Netherlands)

Dr. Martijn Brouwers stepped up next to discuss additional safety outcomes of interest. Treatment discontinuation was significantly higher among participants treated with metformin (41 cases, amounting to 27%) than placebo (19 cases, amounting to 12%; p=0.0002), and accordingly total adverse events occurred significantly more frequently in the metformin-treated group (27% vs. 12%; p<0.001). As expected with metformin, gastrointestinal issues were the most commonly reported adverse events, occurring in 16% of metformin-treated patients versus 3% of those in the placebo group (p<0.001). Additionally, vitamin B12 deficiency occurred significantly more frequently in the metformin-treated group (12%) than the placebo group (5%), translating to a significant elevation in vitamin B12 deficiency risk (HR=2.76; p=0.0094). Dr. Brouwers noted that vitamin B12 deficiency is a known side effect of metformin; the long-term effects of this kind of vitamin B12 deficiency are unclear but if undetected may include aggravation of sensory neuropathy. The REMOVAL data add to a body of evidence that Vitamin B12 should be monitored  during long term metformin therapy: this is rarely done in routine clinical practice.  Finally, there was no increase in hypoglycemia between either treatment arm, and serious adverse events were equivalent between the metformin group (16%) and placebo group (15%).

Conclusions

Peter Rossing, MD (Steno Diabetes Center, Copenhagen, Denmark)

Dr. Peter Rossing provided a helpful summary of the conclusions from each presentation that came before his. He emphasized that REMOVAL is the largest and longest trial of metformin in type 1 diabetes, and that it’s the first RCT to look at an intermediate variable for CV outcomes (this has to do with feasibility, since a traditional CVOT for metformin in type 1 is currently considered too large an investment for diabetes funding agencieswould be a much more massive investment). Despite an underwhelming result on the primary endpoint, Dr. Rossing positioned the statistically significant benefit to tertiary endpoint (maximal cIMT) as an very exciting piece of data, perhaps one to explore further. He reiterated the significant benefits to metformin in terms of weight loss and (modestly) reduced insulin dose. REMOVAL results do not support the assertion in current treatment guidelines that metformin causes a clinically-meaningful improvement in glycemic control, Dr. Rossing explained, but they do suggest that metformin could have a broader role in CV risk management due to the weight loss, LDL-lowering, and slower progression to maximal cIMT. This latter point is will seem like a bold claim to for clinical trial purists, since metformin did not show statistical significance on the primary endpoint of rate of progression to mean cIMT. That said, there have been no cardiovascular outcome trials to date in type 1 diabetes, so we agree with the Commentary in the Lancet Diabetes and Endocrinology that hope is still alive for ’re not ready to count out metformin as a possibility in better management of type 1 diabetes/CV complications.  For one, the generic drug is not going anywhere anytime soon, and it’s a widely-accessible and effective therapy known to be effective in type 2 diabetes. In addition, we’d love to learn more about how elements of trial design could have skewed or muted efficacy findings. For example, a majority of participants were already well-treated on with statins and blood pressure-lowering medications at baseline (which will be hard to avoid in selecting a real-world patient population), and this may have made it more difficult to demonstrate statistically significant benefit on a CV biomarker. Dr. Bill Tamborlane, who chaired the symposium, hinted during Q&A that cIMT may not be the ideal surrogate for CV outcomes, since it didn’t appear after six years of DCCT follow-up and thus may not be sufficiently “timely.” Another point raised during Q&A was that cIMT was already quite “favorable” at baseline, which skews power calculations and maybe have madekes it more challenging to show an simprovement. These ideas are purely speculative right now, and we’ll leave it to the statisticians and other thought leaders to weigh in on REMOVAL at future meetings. Dr. Petrie announced that there will definitely be another symposium dedicated to this type 1 trial at EASD 2017 in Lisbon, Portugal.

Perspective

As commentator, Dr. Naveed Sattar set out to contextualize the REMOVAL results. He set the stage with a discussion of cardiovascular disease in type 1 diabetes. Recent data from a Swedish registry study demonstrates that individuals with type 1 diabetes and no CV risk factors have an 82% greater risk of a CV event than their counterparts in the general population – and yet there are no CV disease primary outcome studies in type 1 diabetes. Against the backdrop of this severe unmet need, the implications of the REMOVAL trial are frustratingly unclear. In Dr. Sattar’s view, a CV benefit is “suggestive but not conclusive” in REMOVAL, making it very unclear whether metformin should be recommended to people with type 1 diabetes to slow the progression of CV disease. On whether further trials are warranted to assess metformin for CV disease in type 1 diabetes, Dr. Sattar was torn. Though metformin is inexpensive and generally safe, making for relatively easy logistics, it seems as though other adjunct therapies (such as SGLT-2 inhibitors and GLP-1 agonists, which have demonstrated cardioprotective effects in type 2 diabetes) are better positioned for this large-scale investment of time and resources (though we’re less optimistic about the future of GLP-1 agonists in type 1 diabetes, given the safety signals in the ADJUNCT trials for liraglutide – though we believe some of this may have been due to trial design and dosing decisions). During Q&A, SGLT-2 inhibitors in particular were highlighted for their potential in type 1 diabetes and the chance that they may confer superior glycemic and CV benefits over metformin. We’re intrigued by this possibility, and we’re closely following Lilly/BI’s and AZ’s and Lexicon’s phase 3 programs for an SGLT-2 in type 1 and/or an SGLT-1/2. Despite REMOVAL’s somewhat underwhelming results, Dr. Sattar closed by reminding the audience of the significance of this undertaking as the single largest and longest randomized CV trial ever to examine a CV parameter of any sort ever in type 1 diabetes. We certainly hope it isn’t the last and to the point raised in Q&A, big funds are needed to invest in this (but that are a pittance compared to funds spent each year on avoidable complications).

Panel Discussion

Dr. Bill Tamborlane (Yale University, New Haven, CT): You talked a lot about LDL changing. I’d like to think about what wasn’t said. What happened to HDL?

Dr. Petrie: We saw very similar changes.

Q: One of the things that determines ability to see changes in cIMT is baseline cIMT. Your patients, I have to say, were extremely well-treated at baseline. To what extent do you think that messed up your power calculation? It’s hard to show improvement in endothelial function, as well, if at baseline you already have good endothelial function.

Dr. Hramiak: That’s a good point, that baseline cIMT was favorable. These patients were actually surprisingly well-treated on statins and blood pressure-lowering medications. Certainly, baseline values were lower than expected, and then rate of progression for mean cIMT was lower than anticipated from data from younger DCCT-EDIC patients and from other conditionsin some prior studies we had looked at. But for maximal cIMT, it was higher.

Comment: Of course, because that’s where the lesion is. I recall another study that had low cIMT at baseline and the group made an incorrect negative conclusion about ezetimibe. You’re hurt by this in REMOVAL, I think.

Q: Are these participants a selected group of survivors? They have 32-33 years of T1D…what fraction of people would have already died of CVD before this time point?

Dr. Colhoun: By definition all cohorts are survivor cohorts, unless you begin the study from birth. If you’re asking if this was a population that was depleted of events – I don’t think so. A more interesting question is why in a population where event rates are fairly substantial do we not see substantially abnormal cIMT.

Q: cIMT is like tree rings, so the effect should persist. Will there be a data registry for these participants to see if the changes in cIMT we see at three years will become meaningful later?

Dr. Petrie: We have approval to follow these patients up in longer term. It will take a long time, but we certainly got that approval. As you say, However, they’re no longer on randomized treatment so we would be looking for a “metabolic memory” effect.

Dr. Julio Rosenstock (UT Southwest, Dallas, TX): I want to commend all the investigators. It takes a lot of work to do these kinds of studies, especially with a retinopathy component. There’s a real need for adjunct therapy in type 1 diabetes. But looking at the results of this trial, I basically see a negative study from all perspectives, not just the primary endpoint. You’re being generous on a potential reduction in CV events based entirely on a tertiary outcome. You’re using the word “modest” too much – two units is absolutely nothing in terms of reduced insulin dose. Yes, metformin is out there and it costs $2 at Walmart. But we need to raise the bar. This study would have been promising 10 years ago, when we didn’t have any other options, but now the SGLT-2 class has potential in type 1 diabetes. I take your word that a CVOT will one day be necessary for type 1. I hope this stimulates potential for a large outcomes trial of SGLT-2 inhibitors in type 1 diabetes, but metformin was totally negative in this trial.

Dr. Colhoun: Raise the bar? What we actually need to do is raise the money. There has never been a CVD outcomes study in the type 1 diabetes population for this reason.

Dr. Rosenstock: I agree with that, but metformin is not the drug. An SGLT-2 inhibitor or GLP-1 agonist would be the drug – certainly not metformin.

Dr. Colhoun: In our paper in Lancet we are certainly not claiming that this is a positive trial. We need a sustained and substantial effort to really drive forward a much higher standard and larger trials with CVD endpoints in type 1 diabetes. We spent a lot of time trying to raise the money to do a CVD outcomes trial this with metformin, and it was extremely challenging. Let’s hope it happens someday. What we have done in this trial is create a very good network to show that cardiovascular trials  with a cardiovascular focus can be done in T1D, which is something people have been skeptical of.

Dr. Tamborlane: In the DCCT after six years of follow-up, cIMT differences were basically zero. This speaks to an issue about choice of surrogate outcome. Carotid thickening may not be timely enough to be able to show an effect.

Dr. Petrie: Ultimately, this is the study we designed after a lot of workshop with JDRF, and this is the outcome we chose. Treatment guidelines in the US and UK say you should use metformin for people who have type 1 diabetes and obesity to try and improve their glycemic control. One message from REMOVAL is that you should not – there’s not much of an effect on glucose control. It’s transient, though it’s statistically significant. So, REMOVAL changes our perspective on metformin for type 1 diabetes, pointing us to maybe investigate more cardiovascular effects.

Q: Did you look at sex differences within the groups? I ask because in the T1D Exchange study, the boys had an improvement but the girls did not (and the boys were more abnormal at baseline so maybe had more room to change).

Dr. Petrie: We haven’t looked into that, but we could do it easily post-hoc.  We need to prioritize a few key post hoc questions so as not to over-interpret the subgroups -  this could be a good one.

Q: From a clinician’s perspective, why was metformin chosen and not an alpha-glucosidase inhibitor?

Dr. Petrie: As you know there is a large CVOT involving alpha-glucosidase inhibitors going on in China right now led by the Oxford group, with results expected reasonably soon. The study includes people with prediabetes and type 2 diabetes.  In my view, tThere’s not enough data on alpha-glucosidase inhibitors in type 1 diabetes to justify this kind of trial. All these hypotheses are interesting and we’d love to see many trials for cardiovascular outcomes associated with different agents in type 1 diabetes – in fact we’d like to see any trial assessing this in type 1 diabetes – but if one trial is to be done it will likely be with an SGLT-2 inhibitor, if phase 3 data in type 1 diabetes is promising and shows adequate safety.  (Editor’s note – we’d love to see a CVOT with multiple therapies.)

Dr. Partha Kar (Portsmouth Hospitals, UK): Based on these results, as clinicians, would you prescribe metformin to your patients with type 1 diabetes at high CV risk?

Dr. Hramiak: Probably not. I’m more convinced that in adolescents who are more insulin-resistant and who have weight issues, we should prescribe metformin because they’ll see more benefit in terms of minimizing insulin dose. Hypoglycemia is a huge limiting factor of insulin, which is why we’re looking for adjunct therapies in the first place. Weight loss is also ideal. But to Dr. Rosenstock’s point: adjunct therapies in type 1 to-date have had trouble proving glycemic benefits beyond 0.2-0.3% A1c reductions, so we’re looking for something other than glycemic benefits. We have to combine outcomes as we did in this trial – it’s not just about glycemia, and it’s not just about CV disease.

Dr. Sattar: What would you do?

Dr. Kar: I already use metformin in the way you describe, so I don’t think REMOVAL will change my practice. Right now, I use metformin in type 1 patients who are overweight, so this just gives me more reassurance. I wasn’t expecting a huge CV benefit per se, and I think cIMT is an ambitious endpoint for a three-year study. But I won’t say this is a completely negative trial, because it confirms my use of metformin in the patients I’ve got.

Dr. Tamborlane: Metformin seems to have some effect, but its borderline. GLP-1 agonists are also being studied in type 1 diabetes. You really get substantial weight loss with these agents which may have an advantage that goes beyond just glycemic control. Another great hope is SGLT-2 inhibitors. It’s been a challenge to find drugs with cardioprotective effects for people with type 1 diabetes, but we’re still hoping that the holy grail is out there somewhere.

Dr. Chaturvedi: Is cIMT a poor outcome since it’s a surrogate? At the moment it is one of the best surrogates we have for cardiovascular endpoints that both predicts endpoints and can be employed easily in large clinical settings. We were not expecting an effect on cardiovascular risk through glycemia lowering alone, we were anticipating pleiotropic effects. Lastly, the effects we saw in the cIMT measures in this trial were actually greater than those shown in the DCCT trial. The DCCT trial saw modest effects on cIMT but quite striking reductions in cardiovascular outcomes ultimately. This is to say that it’s a bit unfair to criticize the endpoint we used. The effects on cIMT we saw here aren’t completely negligible.

Dr. Tamborlane: A lot of these patients were on statins, and a lot were on anti-hypertensive medications. We saw very little use of these agents in the DCCT. It’s hard to parse what the confounding effects of these other drugs may be. I would also suggest that another mistake is calling this study “REMOVAL,” because you didn’t remove anything. Maybe you should have called it the SLOW IT DOWN study.

Q: Did you compare the results between people who were obese vs. non-obese at baseline?

Dr. Petrie: These data are quite fresh, so we haven’t looked into that yet. This could be one to prioritize for post hoc analysis – we will present thesesThere will be more subgroup analyses presented at EASD.

Dr. Calhoun: I take your point that DCCT-EDICT trial did in the end teach us about cardiovascular endpoints, but we can’t realistically wait 30 years to see if diabetes drugs have an impact on cardiovascular disease. The take-home message is that we’re trying to cut short answers to these big questions by doing small studies in type 1 diabetes with surrogate outcomes like cIMT. This will probably not resolve our clinical equipoise sufficiently, but it does tell us something. Perhaps Dr. Rosenstock is actually going to raise the money to do a large CVOT.

Q: The 1,000 mg dose of metformin isn’t particularly big, and it’s a lot less than what was used in DCCT and UKPDS. Is there any plan to do a protocol analysis on dose?

Dr. Petrie: I think we can could do some sort of an analysis like that. We don’t have an estimate of the terribly accurate data on dose of metformin taken by each participant over their time in the study based on tablets distributed and  returned, and we wanted to adopt a pragmatic trial design. An important clinical message: when you give people 2,000 mg of metformin, you’re actually giving them less, because some patients will experience gut rotgastrointestinal adverse effects. We could look at subgroups of patients who tolerate metformin better. If you’re not experiencing side-effects (especially GI), and metformin might improve your CV risk profile and it might control your weight, why not take it?

Dr. Tamborlane: This has been a great discussion with terrific questions – one of the more enjoyable sessions I’ve been at.

Symposium: Inhibition of PCSK9 in Dyslipidemia Patients with Diabetes

Alirocumab and Insulin-Treated Diabetes – Insights from the ODYSSEY DM-INSULIN Study

Lawrence Leiter, MD (University of Toronto, Canada)

Dr. Lawrence Leiter presented results from the ODYSSEY DM-INSULIN trial of Sanofi/Regeneron’s PCSK9 inhibitor Praluent (alirocumab) in patients with diabetes on insulin with LDL-C levels>70 mg/dl despite maximum tolerated statin therapy. The double-blindopen-label, 24 week study randomized 517 patients in a 2:1 fashion to alirocumab 75 mg or placebo – the dose of alirocumab was titrated up to 150 mg at 12 weeks for patients who did not achieve an LDL cholesterol less than 70 mg/dl at week 8that point. Of the 517 patients enrolled in the study, 441 had type 2 diabetes and 76 had type 1 diabetes – the results presented by Dr. Leiter focused only on the type 2 diabetes subgroup. The trial met its primary endpoint by demonstrating a placebo-adjusted 49% reduction in LDL cholesterol with alirocumab (p<0.0001). The average absolute change in LDL cholesterol was 52.9 mg/dl (baseline LDL cholesterol=110 mg/dl). Notably, 80% of participants achieved this reduction with the lower 75 mg dose of alirocumab, while 20% of participants were titrated up to the 150 mg dose at 12 weeks. In terms of the time course of the LDL cholesterol reduction, a massive drop in LDL occurred by week eight, and the LDL largely plateaued after that. Dr. Leiter emphasized that the level of LDL cholesterol reduction observed in this trial is very much in line with previously reported trials from the ODYSSEY program that included both people with diabetes and those without. While we’re certainly glad that the LDL cholesterol-lowering effect does not appear to be attenuated in the context of diabetes, we would’ve loved to see an even greater effect in this population given the high residual CV risk associated with the condition.

  • In addition to LDL cholesterol, alirocumab produced positive results for a number of secondary lipid endpoints: placebo-adjusted, alirocumab was associated with a 39% reduction in non-HDL cholesterol (p<0.0001), a 37% reduction in Apo B (p<0.0001), an 18% reduction in Lp(a) (p<0.0001), a 4% increase in (“good”) HDL cholesterol (p=0.01), a 6% decrease in triglycerides (p=0.0902), a 40% decrease in LDL particle number (nominal p<0.0001), and a 3% decrease in LDL particle size (nominal p<0.0001). 76% of participants in the alirocumab arm achieved an LDL cholesterol <70 mg/dl (vs. 7% in placebo, p<0.0001) and 71% of participants achieved a non-HDL cholesterol <100 mg/dl (vs. 14% in placebo, p<0.0001). Alirocumab did not appear to have an impact on glycemic control or management – there was no difference in A1c, fasting plasma glucose, insulin dose, or diabetes drugs.
  • Dr. Leiter characterized alirocumab as safe and well-tolerated in the study. Serious adverse event rate was actually lower in alirocumab compared to placebo. While a smattering of treatment-emergent adverse events was noted, the overall frequency of individual events in both the alirocumab and the placebo groups were fairly low (<5%). Allergic drug reactions, neurologic events, and increase in ALT were rare in both the alirocumab and the placebo groups. Neurocognitive events were somewhat more frequent in the alirocumab group (occurring in 4 participants [1.4%], compared to 0 in placebo).

ODYSSEY DM Program – Design and Study Populations

Robert Henry, MD (UC San Diego, CA)

Dr. Robert Henry took the stage to present results from ODYSSEY DM-DYSLIPIDEMIA, which compared alirocumab twice-weekly (n=276) vs. usual care (n=137) in adults with type 2 diabetes and mixed dyslipidemia (defined by non-HDL ≥100 mg/dl, triglycerides between 150-500 mg/dl, and atherosclerotic CV disease or at least one other CV risk factor). After 24 weeks of treatment, participants in the alirocumab arm experienced a mean 37% reduction in non-HDL cholesterol vs. a 5% reduction in the placebo arm (p<0.0001), amounting to a treatment difference of 33%. Baseline non-HDL was 155 mg/dl in the Praluent group, 162 mg/dl in the usual care group. Notably, this was the first major clinical trial to use change in non-HDL as its primary endpoint – during his talk on study design earlier in the symposium, Dr. Dirk Müller-Wieland contextualized this decision around an increasing number of professional organizations (the National Lipid Association, the European Society for Cardiology, and others) recommending non-HDL as a target for best practice lipid management. Data on non-HDL from ODYSSEY DM-DYSLIPIDEMIA largely matched results from other trials in the ODYSSEY program. Across five studies, subsets of participants with comorbid diabetes/dyslipidemia experienced ~39% non-HDL lowering with alirocumab vs. ~3% with placebo. Turning to secondary endpoints, Dr. Henry shared that 67% of alirocumab-treated patients achieved a non-HDL goal <100 mg/dl at week 24 vs. 18% of usual care patients (p<0.0001). Similarly, 71% of people in the Praluent arm reached an LDL goal <70 mg/dl vs. 16% of people receiving usual care (p<0.0001). Alirocumab showed statistically significant superiority in lowering LDL (treatment difference 43%; p<0.0001), Apo B (treatment difference 32%; p<0.0001), and Lp(a) (treatment difference 28%; p<0.0001). The PCSK9 inhibitor also increased HDL levels by a 6% greater margin than usual care (p=0.0026). As in ODYSSEY DM-INSULIN, alirocumab therapy had no effect on A1c or fasting plasma glucose. We imagine this piece of data will be important for real-world patients/HCPs given concerns surrounding statin-induced diabetes, and we certainly hope that more providers will now consider PCSK9 inhibitor Praluent for their patients with diabetes who also require a lipid-lowering intervention. We also look to payers to increase reimbursement for products in this class as more compelling evidence accumulates (as it stands, a PCSK9 inhibitor prescription is prohibitively expensive for many). Individually, diabetes and dyslipidemia both exacerbate a patient’s CV risk. Together, they confer pretty bad odds for long-term CV health, and these patients are most in-need of aggressive glucose-lowering and lipid-lowering. CVOT data for alirocumab may further support Praluent for people with diabetes; the ODYSSEY Outcomes trial is expected to complete in February 2018.

Oral Presentations: Obesity Pathogenesis and Treatment – Insights from Mouse Models

PB-718, a Dual GLP-1/Glucagon Receptor Agonist Demonstrates Superior Weight Loss Effect and Ameliorates nonalcoholic steatohepatitis (NASH) in Animal Models

Michael Xu, MD (Pegbio, Suzhou, China)

Dr. Michael Xu presented a range of preclinical data on an investigational, PEGylated GLP-1/glucagon receptor dual agonist – Pegbio’s PB-718. In a mouse model of diabetes (n=6), both low and high doses of the agent were associated with superior weight loss and reduced fasting plasma glucose vs. twice-daily liraglutide. In mice with NASH (n=6), both doses led to more weight loss, a greater reduction in liver weight, and a greater lowering of liver weight/body weight ratio vs. placebo (p<0.0001 for all comparisons). Moreover, there was a dose-dependent reduction in NAS score for NASH severity (p<0.01 for lower dose vs. placebo; p<0.0001 for higher dose vs. placebo), which takes into account steatosis, liver inflammation, fibrosis, and other markers. Dr. Xu also reviewed findings from a preclinical investigation of Pegbio’s drug candidate in monkeys. Consistent with the data from rodent studies, PB-718 demonstrated dose-dependent and significant superiority vs. placebo in stimulating weight loss and improving NAS score. Dr. Xu explained that there are strong mechanistic underpinnings that support the utility of GLP-1/glucagon dual agonists for diabetes, obesity, and NASH. Oxynotomodulin, an endogenous dual agonist of GLP-1/glucagon, has been correlated with decreased food intake, enhanced glucose tolerance, and improved liver health. We’re certainly excited by GLP-1/glucagon dual agonists as a potential new therapy class (competitive landscape here), and we’re happy to see companies investing in these agents for diabetes as well as adjacent indications. NASH, in particular, is a therapeutic area of high unmet need with no FDA-approved medicines to-date, and we also see marked room for improvement in available tools for obesity management.

Questions and Answers

Q: Have you compared your agent with semaglutide (Novo Nordisk’s GLP-1 agonist candidate, in phase 2 for obesity and NASH)?

A: No we have not, but we will be eager to compare our compound to semaglutide in some study. Actually, we were approached by them and tried to figure out how to compare these compounds side-by-side. Of course, semaglutide has not yet been approved.

 

--by Melissa An, Adam Brown, Abigail Dove, Helen Gao, Varun Iyengar, Brian Levine, Payal Marathe, Emily Regier, Maeve Serino, and Kelly Close