May 25-29, 2016; Orlando, FL – Days #3-4 Highlights – Draft

Executive Highlights

Our time in sun-splashed Orlando came to an end over the weekend, though not before we took in the final two days of the American Association of Clinical Endocrinologists 25th Annual Scientific and Clinical Congress. Though the closing agenda featured a broad range of topics beyond diabetes and obesity, those two areas still commanded a good deal of the meeting’s focus, as shown through the top ten highlights below. Highlights included Dr. Darren McGuire’s (UT Southwestern, Dallas, TX) retrospective view on the evolution of cardiovascular outcomes trials for diabetes drugs, Dr. Steve Nissen’s (Cleveland Clinic, Cleveland, OH) thoughts on the mechanism of cardioprotection in EMPA-REG OUTCOME, and Dr. Lance Sloan’s (Texas Institute of Kidney and Endocrine Disorders, Lufkin, TX) musings on unexplained cases of DKA with SGLT-2 inhibitors in type 2 diabetes. On the tech side, we heard commentary from a number of speakers on Medtronic’s 670G/Enlite 3 system (investigators are “very happy” with the real-world accuracy of the sensor; a “pivotal study” of the Enlite 3 sensor is planned in pediatric patients; the current algorithm was the result of much iteration, feedback, and more iteration) along with an update from the charismatic Dr. Ed Damiano (Boston University, MA) on the Bionic Pancreas. For all the details from Days #3-4, please see our Drug and Device Highlights below. And see our Day #1 and Day #2 coverage from last week in case you missed it!

Top Drug Highlights

1. Dr. Darren McGuire (UT Southwestern, Dallas, TX) provided a retrospective view on the evolution of cardiovascular outcomes trials (CVOTs) for diabetes drugs and suggested that the controversial 2008 FDA Guidance is here to stay. Dr. McGuire also suggested that the EMPA-REG OUTCOME results demonstrating a cardioprotective benefit for Jardiance (empagliflozin) will likely prove to be a class effect for SGLT-2 inhibitors. Notably, he also said of Jardiance, “I think the stroke signal is too small to make any clinical conclusions” and that “The PCSK9 data is the most exciting science of my career.”

2. Dr. Steve Nissen (Cleveland Clinic, Cleveland, OH) suggested that the modest reductions in A1c, blood pressure, and weight in EMPA-REG OUTCOME could each have accounted for part of the overall CV risk reduction. A panel discussion at the end of the session touched upon a number of additional potential mechanisms of benefit, including myocardial substrate energetics, anti-arrhythmic effects, and direct effects on the kidney.

3. AACE/ACE unveiled its new clinical practice guidelines (executive summary and algorithm) for the comprehensive medical care of patients with obesity in a symposium chaired by Dr. Timothy Garvey (University of Alabama, Birmingham, AL). The algorithm centers on a staging system for obesity that guides evidence-based treatment goals and therapies. For prediabetes and metabolic syndrome, the guidelines recommend a 10% weight loss goal to prevent type 2 diabetes.

4. Dr. Lance Sloan (Texas Institute of Kidney and Endocrine Disorders, Lufkin, TX) suggested that most unexplained cases of DKA with SGLT-2 inhibitors in type 2 diabetes occur in patients with very high A1cs who should be treated with insulin.

5. Dr. Leo Seman (Boehringer Ingelheim, Ridgefield, CT) presented new data from the BI-funded IntroDia study on how discouraging phrases impact patient resistance to add-on type 2 diabetes therapy.

6. In an engaging and passionate presentation, former US Surgeon General Dr. Regina Benjamin discussed the power of prevention for building a stronger and more stable national health system.

Top Device Highlights

1. Medtronic Diabetes’ Dr. Scott Lee discussed the company’s approach to closing the loop, headlined by valuable insight into how the team conscientiously iterated upon its original MiniMed 670G/Enlite 3 algorithm to improve on early “mixed” returns and news that the company plans to conduct a “pivotal study” of the Enlite 3 sensor in pediatric patients.

2. In a talk reminiscent of his ENDO 2016 presentation, Beta Bionics’ Dr. Edward Damiano stated that the advent of the Bionic Pancreas and other automated insulin delivery devices will inevitably shift the patient and provider focus from diabetes management to device maintenance. Said Dr. Damiano about glucagon: “… [the] development in last 5-10 months has made tremendous progress. You’ll hear more about this over the next few months.” He also raised the question whether alcohol affects access to glucagon (to be determined in a future study by clinical collaborator Dr. Steven Russell).

3. Drs. Satish Garg (University of Colorado, Aurora, CO) and William Tamborlane (Yale University School of Medicine, New Haven, CT) shared two valuable insights from their experience running the pivotal trial of Medtronic’s MiniMed 670G/Enlite 3: (i) that the new sensor has demonstrated impressive real-world accuracy; and (ii) that the system’s human factors still require a bit of fine tuning.

4. AACE’s Dr. George Grunberger delivered critical commentary on the challenges of working with Medicare to establish coverage of CGM – “After several trips from us (AACE) and in spite of all the data and position statements, they told us that it all depends on FDA.”

Top Drug Highlights

1. Dr. Steve Nissen (Cleveland Clinic, Cleveland, OH) suggested that the modest reductions in A1c, blood pressure, and weight in EMPA-REG OUTCOME could each have accounted for part of the overall CV risk reduction. He acknowledged that many other more speculative mechanisms could have contributed as well but preferred to focus on the mechanisms that are more fully understood. Dr. Nissen noted that the landmark studies that failed to find a clear relationship between glucose lowering and macrovascular outcomes all used agents that carried a risk of hypoglycemia, meaning the results may not completely apply to classes like SGLT-2 inhibitors that do not raise hypoglycemia risk. He also described the 4 mmHg reduction in systolic blood pressure in the trial as consequential in the right setting and noted that lowering the rate-pressure product (reducing blood pressure without increasing heart rate) could be very beneficial for patients with ischemic heart disease. Finally, Dr. Nissen emphasized that because most of the weight loss with the Lilly/BI SGLT-2 empagliflozin results from loss of visceral fat, it could very well be linked to CV benefit, contrasting it with previous drugs like sibutramine that have given weight loss medications a bit of a bad name in the cardiology field.  Despite these arguments, Dr. Nissen acknowledged that none of these mechanisms could likely explain the profound reduction in CV death in the study – while he celebrated this result as hugely meaningful for patients, he admitted that he remains completely puzzled as to the mechanism. His take on the heart failure results was similar, though he noted that trials of thiazide diuretics have produced the most analogous results. Most conference speakers we have heard (including speakers at the EASD symposium where the EMPA-REG OUTCOME results were first presented) have suggested that the CV benefits were unlikely to be related to glucose, weight, or blood pressure given how modest the effects on those parameters were. However, in the absence of a clear alternative explanation, we would give Dr. Nissen’s sum-of-its parts hypothesis as much weight as any other.

• A panel discussion at the end of the session touched upon a number of additional potential mechanisms of benefit, including myocardial substrate energetics, anti-arrhythmic effects, and direct effects on the kidney. See the detailed discussion section below for an edited transcript of the discussion.  We will be eager to hear more commentary on this at ADA.  

2. Dr. Darren McGuire (UT Southwestern, Dallas, TX) provided a retrospective view on the evolution of cardiovascular outcomes trials (CVOTs) for diabetes drugs and suggested that the controversial 2008 FDA Guidance is here to stay. He argued that the pre-2008 FDA guidelines for the development of diabetes drugs had a very low threshold, allowing approval based on as little as 250 patient-years of exposure. Dr. McGuire attributed this low bar to the desperate need for additional diabetes therapies when the guidelines were put into place in 1990; the only medications on the market at that time were insulin and sulfonylureas. He suggested that due to the proliferation of diabetes drugs and drug classes in the last 10 years, we now have the “luxury of time” to assess how diabetes drugs impact patients in the long run and consider whether new products are better than existing ones. Indeed, many pharmaceutical executives have recently commented that the prolific diabetes drug landscape has raised the bar for new drugs to the point where modest glucose-lowering is not enough to warrant approval and, just as or more importantly, reimbursement. As a result, many stakeholders are looking for cardiovascular risk reduction, weight loss, or other benefits on top of A1c reduction. It is unlikely, in our view, that any compounds would be approved today that cause hypoglycemia or weight gain. Dr. McGuire expressed excitement over the cardioprotective benefit demonstrated in the EMPA-REG OUTCOME, LEADER, SUSTAIN 6, and IRIS trials and asserted that we’re “getting our money’s worth” from the trials. As an aside, we note that cardioprotection is unlikely to be added to semaglutide’s label (related to SUSTAIN 6) without further studies. Given this proof that it’s possible to demonstrate cardiovascular benefit in these FDA-mandated trials, Dr. McGuire suggested that there’s no need to revisit the guidance, except possibly to add heart failure as an endpoint. Heart failure may shape up to be a differentiating endpoint for many diabetes drugs – AZ’s DPP-4 inhibitor Onglyza (saxagliptin) showed a negative impact on heart failure while Lilly/BI’s SGLT-2 inhibitor Jardiance (empagliflozin) demonstrated a positive impact. As we understand it, this might require greater standardization of how heart failure is defined and assessed in clinical trials.

• Dr. McGuire argued that many of the initial concerns regarding the 2008 FDA guidance ended up being unfounded. While critics at the time argued that the diabetes field didn’t have the money, infrastructure, or patient recruitment capabilities to conduct these large-scale trials, Dr. McGuire pointed out that a quarter of a million patients have now been enrolled or plan to enroll in these trials. In addition, he emphasized that a “non-inferior” result in these trials is a positive outcome, as improving microvascular outcomes via glucose-lowering without increasing macrovascular risk is a worthy goal in and of itself. Given the recent spate of CVOTs demonstrating superiority, this is a timely reminder as we consider those that are positive for non-inferiority, such as Intarcia’s FREEDOM-CVO for ITCA 650 that may be positive for superiority if studied for a longer period of time. However, while we agree that demonstrating CV safety is a worthwhile goal, we do not necessarily agree that it is always the best use of limited research dollars for drugs with no signal of increased risk.  To boot, we certainly believe that the rise of required often pre-approval CVOTs has encouraged a range of companies to move away from investing in diabetes, including BMS, Genentech, Novartis, and others. Given that nearly half of all patients are not at their glycemic targets and that future doctors are discouraged from entering the field, we would hardly say that field doesn’t need further investment in innovation.
• As the architect of the FDA’s 2008 CV Guidance, Dr. Steve Nissen argued in a subsequent talk that studies like EMPA-REG OUTCOME and LEADER were always the goal of this requirement. He stated that while safety was the immediate impetus for the requirement, his ultimate goal was always to encourage development of glucose-lowering therapies that reduce the burden of macrovascular disease. He applauded the “courageous” decision by the FDA Advisory Committee to support the very controversial guidelines and suggested that his position had been vindicated. While we would support the studies that show safety, we would also point out that the guidelines discourage small, entrepreneurial companies from entering the fray. “We’ll jut do oncology ….”
• We would argue that the implications of recent positive CVOT results for the broader requirement are a bit more nuanced. We acknowledge that the diabetes field has derived more value from these trials than some had anticipated, and it is quite possible that the field would never have known about the CV benefits of products like Lilly/BI’s Jardiance and Novo Nordisk’s Victoza (liraglutide) in the absence of an FDA requirement to conduct an outcomes trial (although we suspect the companies would’ve done CVOTs to drive demand, it is impossible to prove this). However, we still believe it would make more sense for these decisions to be made on a case-by-case basis; our ideal situation would be for the FDA to require a CV safety trial when there is concern that a particular agent may carry increased risk and for companies to have the right incentives to voluntarily conduct a superiority trial when there is clear potential for benefit. Dr. John Jenkins of the FDA suggested in 2014 that the field should have a broad discussion of this at some sort of advisory meeting, but this has yet to be announced. From that meeting, according to Dr. Jenkins: “… it may be time to revisit the basis of cardiovascular studies for diabetes drugs …”
• Dr. McGuire also suggested that the EMPA-REG OUTCOME results demonstrating a cardioprotective benefit for Jardiance (empagliflozin) will likely prove to be a class effect for SGLT-2 inhibitors. He pointed to early meta-analyses of J&J’s Invokana (canagliflozin) and AZ’s Farxiga (dapagliflozin) that demonstrated similarly large effects on mortality, though he acknowledged that it is possible that the meta-analyses’ results were artifacts of the small event number.

3. AACE/ACE unveiled its new clinical practice guidelines (executive summary and algorithm) for the comprehensive medical care of patients with obesity in a symposium chaired by Dr. Timothy Garvey (University of Alabama, Birmingham, AL). These evidence-based guidelines were developed to facilitate high-quality care of patients with obesity for clinical decision-making that addresses the real-world medical care of patients with obesity, including screening, diagnosis, evaluation, therapy selection, treatment goals, and individualized care. The algorithm centers on an obesity staging system that guides treatment goals and therapies with screening suggested for patients presenting with a BMI over 25 kg/m² (or over 23 kg/m² for some ethnicities including South Asian, Southeast Asian, and East Asian) and/or with weight-related conditions that could be improved with weight loss. Diagnosis and staging of obesity is based on (i) an evaluation of the patient’s medical history, physical examination, and obesity history, (ii) anthropometric diagnosis via BMI and waist circumference, and (iii) clinical diagnosis based on BMI and presence of obesity-related complications. Treatment recommendations are dependent on the stage of obesity, but AACE/ACE recommend lifestyle modification for all individuals, with pharmacotherapy as an option when warranted. See below for our detailed discussion and commentary on how the new algorithm tackles obesity pharmacotherapy, complications-based treatment goals, lifestyle therapy, and more.

• The guidelines utilize a complications-centered approach to the staging obesity, looking beyond BMI and allowing for the enhancement of the overall health of patients. “Stage 0” includes patients who are overweight (BMI 25-29.9) or have obesity (BMI ≥30) with no complications. “Stage 1” includes patients with BMI ≥25 with one or more mild-to-moderate weight-related complications or who may be treated with moderate weight loss and “Stage 2” includes patients with BMI ≥25 with at least one severe complication or require more aggressive weight loss for effective treatment. Individuals with BMI <25 kg/m² (or <23 kg/m² for certain ethnicities) and waist circumference below regional/ethnic cutoffs are considered normal weight, and the guidelines recommend primary prevention of obesity in these patients.
  
  • For Stage 0, AACE/ACE recommends secondary prevention through lifestyle modification, though pharmacotherapy is an option if lifestyle recommendations are not effective. In both Stage 1 and Stage 2, the guidelines recommend treatment focused on achieving weight loss to improve existing complications and prevent further ones from developing, utilizing lifestyle therapy for all patients and suggest pharmacotherapy for those with a BMI ≥27 kg/m². The guidelines for stage 2 also recommend considering bariatric surgery in patients with BMI ≥35 kg/m². Follow-up guidance makes clear that weight loss plateau is expected and complications should be re-evaluated at this point to determine if treatment should be intensified.
• We applaud the guidelines for stratifying different levels of obesity into stages based on associated complications and providing comprehensive evidence-based guidance for providers on how to optimize use of the various obesity therapies within their arsenal. We hope these guidelines can improve uptake of obesity treatments and give providers much needed clarity on their appropriate usage. We look forward to understanding more how clear PCPs and others with little time with patients perceive the guidelines. We also urge payers to take note of these guidelines and implement greater reimbursement of obesity pharmacotherapies, a major ongoing barrier for patients gaining access access to obesity care. More data on how much conversation to diabetes these guidelines help prevent would be useful over time.

4. Dr. Lance Sloan (Texas Institute of Kidney and Endocrine Disorders, Lufkin, TX) suggested that most unexplained cases of DKA with SGLT-2 inhibitors in type 2 diabetes occur in patients with very high A1cs who should be treated with insulin. He noted that the reported cases of SGLT-2 inhibitor-associated DKA in type 2 diabetes seem to fall into four groups: (i) people who were not really in DKA; (ii) people who really have type 1 diabetes or LADA; (iii) people with contributing factors like illness or surgery; and (iv) “the group we’re really interested in” (presumably due to its relatively large size) – people with type 2 diabetes with no contributing factors. He suggested, based on his experience and a review of recent posters, that most patients in the latter group have very poor glycemic control (A1c >12%) and go into DKA after beginning treatment with an SGLT-2 inhibitor and no insulin. His hypothesis is that these patients have low (or presumably no) beta cell function, leading to decreased insulin secretion and very high blood glucose levels, which in turn lead to glucose toxicity that further worsens beta cell function, and to intravascular volume contraction that leads to dehydration. Adding an SGLT-2 inhibitor in this setting can worsen some of these defects by lowering insulin secretion (through lowering glucose), increasing glucagon levels, and promoting diuresis. The patient can then enter a vicious cycle of diuresis leading to further intravascular volume contraction and release of stress hormones, which in turn leads to higher blood glucose, further diuresis, etc. All of these factors together constitute a perfect storm for DKA. Dr. Sloan believes the same basic process can apply at lower A1cs as well, but typically only in the presence of a predisposing factor. His solution? “Practice good clinical medicine.” He emphasized that any patient with an A1c >10% needs insulin at least initially to break glucose toxicity, and treatment can be adjusted once glycemic control is improved. We found it fairly shocking that so many people with A1cs over 12% would receive a recommendation that did not include insulin.

• Dr. Sloan suggested that SGLT-2 inhibitors can be useful in patients with type 1 diabetes and obesity but advised against using them in patients with classic type 1 diabetes. He believes using an SGLT-2 inhibitor in patients with type 1 diabetes taking < 0.5 U/kg (50 units) of insulin per day carries too high a risk of DKA and that weight loss is the clearest reason to use the class off-label in this population. He acknowledged that other factors like decreased variability (particularly hyperglycemia) and possible cardio- and renal protection contribute to the class’ appeal even for normal weight patients but stressed that safety is paramount. Larger phase 3 trials should provide more insight about the risks and benefits of SGLT-2 inhibitors in the broad type 1 diabetes population. We do believe that the benefits – particularly reduced variability – of these drugs may outweigh the risks for some patients with type 1 diabetes who do not need to lose weight or that have too much hyperglycemia, though it will be crucial that patients and providers are educated about the risk of DKA and how to mitigate it. Recent work from dQ&A (contact Richard.wood@d-qa.com) showed that very few patients had access to a ketone meter or understood what appropriate ketone results should be. As discussed in our ADA Postgrad 2016 report, the diabetes market research company dQ&A was particularly struck in a recent panel survey (they survey thousands of patients each quarter) by the lack of preparedness for DKA among patients with type 2 diabetes, even those taking an SGLT-2. Surprisingly, most patients had not talked to their doctor about it, and few own a ketone testing kit or were confident that they could recognize the symptoms of DKA. It is clear that there is much work that needs to be done to increase patient education about DKA and ketone testing and to increase access to ketone testing kits, but we remain optimistic that this is addressable. There’s more detail on dQ&A’s questions and answers about DKA and ketone testing here or you can reach out directly to Richard Wood as shown above.

5. Dr. Leo Seman (Boehringer Ingelheim, Ridgefield, CT) presented new data from the BI-funded IntroDia study on how discouraging phrases impact patient resistance to add-on type 2 diabetes therapy. This segment of the study involved survey data from 4,235 patients adding an additional oral medication or a combination oral medication to an existing single or dual oral therapy for type 2 diabetes. The overarching IntroDia study also involved surveys of physicians (n=6,753) and of patients at the time of diagnosis (n=5,904). Patients who were on three or more medications, adding an injectable medication, or with a diabetes duration over five years were excluded. Patients were surveyed on their experience in the add-on moment (with a questionnaire asking whether 24 specific conversation elements were involved in the conversation), their perceived communication quality, and their self-reported outcomes. The study found that 20% of patients “bargain” with their provider during this conversation to delay therapy intensification. These patients gave the interaction poorer perceived communication quality scores than patients who did not bargain (p<0.001). Importantly, a larger percentage of patients who did bargain recalled “discouraging” (as opposed to “encouraging” or “cooperative”) conversation elements in the interaction than those who did not bargain – see the table below for the full results. We hope training for healthcare professionals can begin to incorporate these insights to promote more encouraging and collaborative conversations with patients about therapy intensification – we were quite shocked at some of the conversational elements that were presented as neutral. In addition, we hope the characteristics of the patients who bargained that emerged from this study can help identify which patients might need extra attention, support, and education when considering therapy intensification.

• The study also found large regional variation in the percentage of patients who did or did not bargain during this conversation, ranging from 53% of patients in South Korea resisting additional therapy to just 6% of patients in Australia doing the same, which we found quite fascinating. Characteristics of patients who bargained also differed from those who didn’t bargain – those that did tended to experience greater diabetes distress in terms of emotional burden (p<0.001) and regimen-related burden (p<0.001), had poorer diets (p<0.001), and were less adherent to their medications (p<0.001) than those who did not bargain. These results demonstrate significant, important differences between patients who readily accept type 2 diabetes therapy intensification and those who are initially more resistant to intensification.

Table 1: Percentage of patients recalling discouraging conversation elements in add-on therapy discussion

6. In an engaging and passionate presentation, former US Surgeon General Dr. Regina Benjamin discussed the power of prevention for building a stronger and more stable national health system. According to Dr. Benjamin, giving Americans health coverage is just a small step towards tackling health disparities in the US as it does not solve the social determinants of health: poverty, smoking, education attainment, race/ethnicity, alcohol consumption, and other very troubling issues. To address these determinants and improve the health of the nation, Dr. Benjamin dedicated much of her time as Surgeon General to advocating for prevention – or as she characterized it, a “culture of health”, as we often hear characterized by the Robert Wood Johnson Foundation. Indeed, in 2011 she and 17 cabinet-level members met and produced the first-ever National Prevention Strategy with the goal of “increasing the number of Americans who are healthy at every stage of life.” She noted that the lack of prevention has taken a substantial toll in recent years, with ~50% of all adults having at least one chronic condition caused by four modifiable risk factors: lack of physical activity, poor nutrition, tobacco use, and excessive alcohol use. In Dr. Benjamin’s view, one of the most troubling outcomes of poor prevention is the skyrocketing rates of obesity, which she noted are disproportionately high among black, Hispanic, and Native American populations. During her time as Surgeon General, she aimed to reduce the barriers to physical activity through multiple programs such as the Every Body Walk campaign and Zumba on the Plaza. Dr. Benjamin ended her presentation by calling for doctors to lead by example and encourage others to be healthy: “you have to find your healthcare joy and help your patients find theirs … our patients trust us just because we are doctors and we have the responsibility of being good leaders.” We admire Dr. Benjamin for her passion for health equity and prevention, and are happy to see that current Surgeon General Dr. Vivek Murthy has continued to champion the National Prevention Strategy.

Top Device Highlights

1. Dr. Scott Lee’s (Medical Director, Medtronic Diabetes, Northridge, CA) overview of Medtronic’s approach to closing the loop was headlined by valuable insight into how the team conscientiously iterated upon its original MiniMed 670G/Enlite 3 algorithm to improve on early “mixed” returns and news that the company plans to conduct a “pivotal study” of the Enlite 3 sensor in pediatric patients. The crux of the presentation summarized the tremendous amount that was learned from early feasibility studies of the MiniMed 670G/Enlite 3, with Dr. Lee noting that the company came out of its early trials with two big insights: (i) “the algorithm was WAY too conservative” during the day; and (ii) “patients actually wanted some degree of control” vs. full automation. The first lesson was not much of a surprise, as Dr. Lee explained that the company deliberately erred on the side of safety (e.g., under-delivering insulin to avoid hypoglycemia), though the latter point came as a reality check – that some patients are not going to feel comfortable entirely relinquishing their diabetes control to a device - we imagine this will change as comfort builds but are glad the company recognized this as we believe many patients will feel this way (whether or not it is apt). To address this feedback, Dr. Lee summarized four strategic choices that have enabled Medtronic to improve the system: (i) lowering the “correction target” to 150 mg/dl from 180 mg/dl when patients choose to override the system and give insulin themselves; (ii) reconfiguring the algorithm to make adjustments more quickly and dose insulin more aggressively; (iii) allowing patients to determine their own insulin-to-carbohydrate ratio rather than having the system calculate this automatically; and (iv) keeping the overall algorithm set point at 120 mg/dl. Dr. Lee stressed that the sum of these changes have allowed Medtronic to significantly tighten daytime glycemic control and account for patients inevitably fiddling with the system during waking hours. Ultimately, he did not share any new data but his words did offer confidence that 670G pivotal results at ADA would meet our expectations – meaningful improvements in hypoglycemia, time-in-range, and potentially mean glucose.

• Though it was not mentioned during his talk, Dr. Lee included a slide showing that a pediatric “pivotal study” of the Enlite 3 is on Medtronic’s radar. This is the first we have heard of such plans, though there was no timing listed. The study will surely be critical in expanding the company’s hybrid closed-loop and Guardian Connect indications below age 14. We also expect a pediatric indication to present another early competitive barrier to entry, as smaller companies will likely pursue adult approval first. As a reminder, the Enlite 3 has only been tested in children > 14 years to date.

2. In a talk reminiscent of his ENDO 2016 presentation, Beta Bionics’ Dr. Edward Damiano stated that the advent of the Bionic Pancreas and other automated insulin delivery devices will inevitably shift the patient and provider focus from diabetes management to device maintenance. According to Dr. Damiano, while the Bionic Pancreas will greatly ease and improve diabetes management, it is not a replacement organ and will still require the burden of regular device maintenance – “the hardest part will be getting people to keep wearing these things,” he stated. Indeed, pump adoption has remained low among the insulin-dependent population, with penetration only reaching about one-third of type 1s by most estimates. Beyond cost and reimbursement issues, there are many patients who simply don’t want the added hassle and complexity of cannula insertion, pump failures, additional cost, and wearing something on the body. Dr. Damiano argued that Bionic Pancreas adoption won’t be as much of a problem for patients accustomed to pump therapy and/or MDI due to their frustrations with those standard-of-care therapies coupled with the proven superiority and enhanced convenience of the Bionic Pancreas. However, he believes that patients with new diagnoses will likely be more resistant to wearing and maintaining the device 24/7 – “people who have lived with type 1 are going to love it, but people who are newly diagnosed will hate it.” To that end, Dr. Damiano suggested that the role of the endocrinologist will evolve to focus largely on encouraging and supporting technology use in patients with diabetes. This was a fascinating perspective, as many would argue the opposite: longstanding type 1s on MDI may resist the Bionic Pancreas (“I’m set in my ways! What I have works fine!”), while those newer to diabetes may embrace technology. In one sense automated insulin delivery will ease providers’ lives (e.g., no need to titrate basal rates and insulin-to-carbohydrate ratios), but on the other hand, they may need to become tech support gurus and masters of appealing payer decisions. In an ideal world, diabetes technology will free healthcare providers to focus on supporting adoption and the emotional and psychological stress of living with the disease. 

• See the detailed discussion section below for the full transcript of the discussion.

3. Drs. Satish Garg (University of Colorado, Aurora, CO) and William Tamborlane (Yale University School of Medicine, New Haven, CT) shared two valuable insights from their experience running the pivotal trial of Medtronic’s MiniMed 670G/Enlite 3: (i) that the new sensor has demonstrated impressive real-world accuracy; and (ii) that the system’s human factors still require a bit of fine tuning. Commentary on CGM performance came during Q&A in response to an audience question that asked panelists to identify the “best CGM” on the market. Both Drs. Garg and Tamborlane hedged their answers and stressed that “all sensors have come a long ways” though did acknowledge that Dexcom has traditionally been a bit stronger on the accuracy front (an understatement from the view of the US, though the 670G has been perceived as more accurate in tests outside the US). Dr. Tamborlane then alluded to his recent experience with the Enlite 3 from Medtronic’s pivotal study, noting that the company is “catching up” and that his clinical staff has been “very happy” with the new sensor’s performance. As a reminder, data from ATTD showed a MARD of 11% vs. YSI for the Enlite 3 in a pre-pivotal study – that’s far better than the original Enlite on paper and it’s even better to hear that the real-world experience bears this out. More critically, Dr. Garg suggested that there remains room to improve the human factors of the hybrid closed-loop design, anecdotally reporting that nurses “needed two-to-four weeks of hand-holding” to become comfortable with the system. This is to be expected for a first-gen device, and we assume the 670G will indeed be a step change in setup complexity vs. the simple 530G and even simpler predecessors. Dr. Garg’s broader point was well taken: there is still significant room to improve human factors in even the most cutting-edge technologies, and widespread adoption of the 670G out of the gate may be too much to expect. He urged industry not to rest in its effort to ensure that devices fit seamlessly into patients’ and providers’ lives and stressed that even the best diabetes devices will only be effective if they are easy to use. We believe “presentation” is critically important – given that over time, devices will undoubtedly become easier to use, patients and providers may need to be reminded of the importance of early feedback and how necessary that is to progress.

• Dr. Garg also shared fascinating data (new to us) documenting the tremendous real-world hypoglycemia benefit of threshold suspend and resulting (theoretical) cost savings for a payer. He presented aggregate numbers from 54,268 US threshold-suspend users (4,449,952 days of data; 2,607,316 threshold suspend events) that showed 67% and 36% reductions in spent time below 50 mg/dl and 80 mg/dl, respectively, when threshold suspend was turned on vs. turned off. A subsequent slide translated these results into a theoretical cost savings for payers under a few assumptions: (i) patients not using threshold suspend accrue an additional 0.87 severe hypoglycemia events per year requiring medical assistance; (ii) a single emergency visit costs an estimated $1,387 (this is a low estimate from our view); and (iii) a single in-patient visit costs an estimated $17,564. The slide suggested that the hypoglycemia difference equates to savings of ~$1,200-$15,000 per patient per year, and we have to think that such an analysis would be persuasive in front of payers. Indeed, we imagine this was particularly convincing in front of UnitedHealthcare, who selected Medtronic as its “preferred” provider of insulin pumps in May. The big question, of course, is whether standalone CGM can offer similar reductions in hypoglycemia. We look forward to seeing what Dexcom’s DiaMond study shows at ADA. 
• Both Dr. Garg and Dr. Tamborlane’s presentations came during an impressively well-attended 5 AM technology symposium that was at 80% capacity despite the early hour. The event was just as highly attended – if not more – than some of the evening symposia at this year’s AACE meeting, a testament to endocrinologists’ interest in getting up to speed on the latest in diabetes technology … and a testament to how quickly this field is changing.

4. AACE’s Dr. George Grunberger delivered commentary on the challenges of working with Medicare to establish coverage of CGM – “After several trips from us (AACE) and in spite of all the data and position statements, they told us that it all depends on FDA.” He expressed a great deal of frustration and generally echoed much of the exasperation we’ve heard from other advocates. It continues to strikes us how difficult it is to have the diabetes voice heard on Capital Hill despite the effort of multiple stakeholders (patient organizations, providers associations, industry groups, etc.) that have been working to convince the organization of the profound logic of CGM coverage. Dr. Grunberger summarized that Medicare’s primary objection to covering CGM remains the adjunctive labeling (i.e., that treatment decisions be based on BGM readings rather than CGM readings) though we’re hopeful that Dexcom is close to overcoming this objection. As a reminder, the FDA is holding a July 21 advisory panel meeting to discuss a non-adjunctive, insulin-dosing label claim for Dexcom CGM. While we are disappointed that the FDA is requesting this advisory committee (which does not seem like a prudent use of resources), we are hopeful that this is a question of “when,” not “if.” Dr. Grunberger’s commentary reinforced that the biggest hurdle in working with CMS is not a lack of clinical data but the bureaucratic red tape muddling the process, and we got the sense that many attendees left the session flummoxed by just how challenging the political angle to this problem is – how to deal with an agency that isn’t responding to reason? We salute Dr. Grunberger and collaborators for continuing to drive awareness in Washington and wonder if there is any chance of the Medicare CGM Access Act of 2015 passing (following its reintroduction in the Senate and House of Representatives in late March) – we certainly assume the July 21 meeting would have to go positively in order to have a hope of this passing.

Detailed Discussion and Commentary

Diabetes Update 2016

Panel Discussion

Q: Do you think it’s possible that SGLT-2 inhibitors might work like eplerenone or Aldactone?

Dr. Steve Nissen (Cleveland Clinic, Cleveland, OH): The paradox is that if anything, it looks like they result in RAS activation rather than depression. I don’t think that’s likely the mechanism but there’s a lot we don’t know. As these trials come in, we’ll need careful sub-studies to understand the biology. I’m still very puzzled by the magnitude of the heart failure and CV death outcomes.

Dr. Darren McGuire (UT Southwestern Medical Center, Dallas, TX): That idea resonates with me as well. I think it might be modifying the renin-angiotensin-aldosterone system long-term. If the mechanism of blood pressure reduction is diuresis, as renal function wanes that should diminish and the opposite occurs. There’s greater blood pressure reduction in patients with kidney impairment. The drop in urine albumin is immediate, sometimes before diuresis has started. There’s something going on at the kidney level. Sodium goes with glucose and chloride goes with sodium, and it immediately shuts down hyperfiltration. If you believe all of this, it may be that you need a certain level of cardiac and kidney dysfunction to see the benefits. Lots of people want to paint the world with this drug. There’s reason to believe that the treatment benefit is related to underlying cardiovascular and kidney dysfunction.

Q: On the increase in stroke, have you looked at what type of strokes occurred? Are you careful with your current patients who’ve had a stroke?

Dr. McGuire: I think the stroke signal is too small to make any clinical conclusions. If anything, a drug that reduces blood pressure should reduce stroke. If you go back to ADVANCE and ACCORD, you don’t see separation of event curves for stroke with blood pressure lowering for three years. These studies may be too short. We went in thinking the moment you lower blood pressure you should affect stroke risk, but in the diabetic population that’s not panning out. MI goes in the right direction, stroke goes in the wrong direction but it’s not significant. My conclusion is that it doesn’t affect those outcomes.

Dr. Nissen: There’s lots of uncertainty. Keep in mind studies like EMPA-REG OUTCOME are long enough that blood pressure reduction might have affected stroke. We know that if we lower blood pressure in the wrong patients, we do have a cerebrovascular downside. I think it’s really interesting. It may be about the specific nature of the population.

Dr. Robert Henry (UCSD, San Diego, CA): The evidence of benefits occurred very quickly. The difference in standard MACE occurred within weeks to a month. Hospitalization for heart failure showed up in the first measurement with a huge split. Which of the things you mentioned worked so fast that you got a signal at the first measurement?

Dr. Nissen: I’m as puzzled as you. At least on the heart failure side, if you reduce intravascular volume, as they showed in the thiazide studies, the relative risk for heart failure is the largest of the benefits. Heart failure is a pretty bad thing in diabetes and can lead to mortality. It may be somewhat driven by that. I think the comments about what’s going on in the kidney are relevant. There may be a lot there that we haven’t sorted out.

Dr. Henry: At ADA there will be a perspective on the issue of energetics, changes in energy. Nobody knows energy better than a cardiologist. It takes lots of energy to pump the heart.

Dr. McGuire: Lots of things are being speculated and pursued. One is the effect on myocardial substrate energetics. There are SGLT-1 receptors on cardiac myocytes. There’s some evidence that if you inhibit SGLT-2 in the kidney, SGLT-1 is upregulated in the kidney. It could also be enhanced in the heart. If so, that could cause a shift toward stress ischemic glucose metabolism. If you look at the separation of the death curves, it looks like an anti-arrhythmic effect. EMPA-REG OUTCOME defined sudden cardiac death as being alive and well and within two hours being found dead. Typically in trials they want to give you the benefit of bedtime – you go to bed well and wake up not. In EMPA-REG OUTCOME that was not the case. There was a two-fold reduction in what I would call sudden cardiac death – that’s anti-arrhythmic. The only thing I’m aware of that had that rapid separation is the mineralocorticoid receptor antagonists in low-ejection fraction heart failure. It looks like the aldosterone antagonists.

Dr. Nissen: One of my favorite phrases is that the road to hell is paved with biological plausibility. We get a result and then we churn and churn to come up with biologically plausible explanations. When it’s done after the fact, I always worry we’re wrong. That’s why I didn’t go into that. I’m aware of those hypotheses but I want more meat on the bones. So I stuck with the things we know about, but I’m not sure you’re wrong.

Dr. Robert Eckel (University of Colorado, Aurora, CO): The broad application of SGLT-2 inhibitors to people without cardiovascular disease is premature. The population studied in EMPA-REG OUTCOME was very selected. We need to wait for more data. Substrate oxidation in the heart is mostly fatty acid-driven. The whole issue of myocardial substrate preference is interesting.

Dr. Henry: Are the benefits in the kidney because of the heart or are they independent effects?

Dr. McGuire: I think they’re independent. I was recently taught by a nephrologist that the kidney is not one big nephron. I had the sense that all nephrons would be similarly diseased, but at an eGFR of 40 there are some that are completely normal, some limping along, and some dead. If we can favorably affect the hyperfiltering sick nephrons selectively, we could get acute effects that are independent of the cardiac benefits. I think the effects are way too early to attribute to glucose.

Dr. Lance Sloan (Texas Institute of Kidney and Endocrine Disorders, Lufkin, TX): As a nephrologist, I just want to say that the EMPA-REG OUTCOME trial would have happened regardless of the FDA because we predicted this. There are a number of mechanisms that would make you expect an improvement in kidney outcomes. Some are hemodynamic and occur rapidly, and there are metabolic things like lowering sugar that are also important but we just won’t see till later on. This is a very important class of medications, and these things were predicted. It was not unexpected.

Dr. McGuire: I live in UT Southwestern where we have lived LDL for 30-40 years. I’ve concluded it’s completely useless. It’s the terminal waste product of VLDL processing. If you lower LDL by clearing it, is there a level that’s too low?

Dr. Pam Taub (UCSD, San Diego, CA): It’s an interesting concept. If you look back 400-500 years, LDL in the hunter-gatherer population was pretty low, in the 30s. As we became industrialized, LDL crept up. What is a truly normal LDL level? The main concern about lowering it too much is cognitive effects. People have speculated that cholesterol is an important building block for neurocognitive function. That hasn’t played out in the PCSK9 trials. In subgroup analyses, they looked at people with two LDL values <50 mg/dl to see if there was a higher incidence of neurocognitive deficits or any adverse effects, and they haven’t seen anything. You can drive it low and not have significant effects. The other evidence comes from the Dallas Heart Study from people with natural PCSK9 mutations. We’re not seeing cognitive deficits or other adverse events there either. I say drive as low as you can safely go.

Dr. McGuire: The PCSK9 data is the most exciting science of my career. We were fortunate to have a single person who was a homozygote knockout. She had a master’s degree and was an aerobics instructor with three kids. So she was mentally intact, neuromuscularly intact, and reproductively intact. Her LDL was 7-11 mg/dl. LDL to my knowledge has no role in brain synthesis. It’s garbage. As long as you lower LDL by clearing it, I don’t perceive any LDL that’s too low. We had intense conversations with the FDA about PCSK9 inhibitors. Amgen decided to hold the dose and Regeneron decided to give it. Fortunately we have that data, and people do well.

Dr. Taub: One controversy is that there are lots of neurocognitive effects with statins. I think that’s possibly mitochondrially mediated. Do you think the side effects with statins are real?

Dr. McGuire: I’m not convinced loss of cognition is a true side effect. I thought the FDA was premature to put it in the label. I think the diabetes risk is real but I’m not convinced the cognitive effects are. Statins are much dirtier; they do a lot other than just lower LDL. PCSK9s are precise.

Dr. Nissen: I get asked by patients all the time how low is too low. I say if you go below zero, I begin to worry.

AACE Comprehensive Clinical Practice Guidelines for Medical Care of Patients with Obesity

Timothy Garvey, MD (University of Alabama, Birmingham, AL), Jeffrey Mechanick, MD (Mt. Sinai Hospital, New York, NY), Karl Nadolsky, DO (Diabetes, Obesity & Metabolic Institute, Bethesda, MD), Ania Jastreboff, MD (Yale University, New Haven, CT), Daniel Hurley, MD (Mayo School of Medicine, Rochester, MN)

AACE/ACE unveiled its new clinical practice guidelines (executive summary and algorithm) for the comprehensive medical care of patients with obesity in a symposium chaired by Dr. Timothy Garvey (University of Alabama, Birmingham, AL). These evidence-based guidelines were developed to facilitate high-quality care of patients with obesity for clinical decision-making that addresses the real-world medical care of patients with obesity, including screening, diagnosis, evaluation, therapy selection, treatment goals, and individualized care. The algorithm centers on an obesity staging system that guides treatment goals and therapies with screening suggested for patients presenting with a BMI over 25 kg/m² (or over 23 kg/m² for some ethnicities including South Asian, Southeast Asian, and East Asian) and/or with weight-related conditions that could be improved with weight loss. Diagnosis and staging of obesity is based on (i) an evaluation of the patient’s medical history, physical examination, and obesity history, (ii) anthropometric diagnosis via BMI and waist circumference, and (iii) clinical diagnosis based on BMI and presence of obesity-related complications. Treatment recommendations are dependent on the stage of obesity, but AACE/ACE recommend lifestyle modification for all individuals, with pharmacotherapy as an option when warranted. See below for our detailed discussion and commentary on how the new algorithm tackles obesity pharmacotherapy, complications-based treatment goals, lifestyle therapy, and more.

• The guidelines utilize a complications-centered approach to the staging obesity, looking beyond BMI and allowing for the enhancement of the overall health of patients. “Stage 0” includes patients who are overweight (BMI 25-29.9) or have obesity (BMI ≥30) with no complications. “Stage 1” includes patients with BMI ≥25 with one or more mild-to-moderate weight-related complications or who may be treated with moderate weight loss and “Stage 2” includes patients with BMI ≥25 with at least one severe complication or require more aggressive weight loss for effective treatment. Individuals with BMI <25 kg/m² (or <23 kg/m² for certain ethnicities) and waist circumference below regional/ethnic cutoffs are considered normal weight, and the guidelines recommend primary prevention of obesity in these patients.
  
  • For Stage 0, AACE/ACE recommends secondary prevention through lifestyle modification, though pharmacotherapy is an option if lifestyle recommendations are not effective. In both Stage 1 and Stage 2, the guidelines recommend treatment focused on achieving weight loss to improve existing complications and prevent further ones from developing, utilizing lifestyle therapy for all patients and suggest pharmacotherapy for those with a BMI ≥27 kg/m². The guidelines for stage 2 also recommend considering bariatric surgery in patients with BMI ≥35 kg/m². Follow-up guidance makes clear that weight loss plateau is expected and complications should be re-evaluated at this point to determine if treatment should be intensified.
• The AACE/ACE algorithm includes a framework for when to initiate pharmacotherapy and substantial guidance on individualization based on the presence of specific complications. The guidelines recommend lifestyle therapy (meal plans, physical activity, and behavior modification) in patients with no complications or mild-to-moderate complications and recommend pharmacotherapy as an adjunct in the following cases: (i) patients with progressive weight gain or no improvement in complications with lifestyle therapy; (ii) patients who are experiencing weight regain with lifestyle therapy after initially having success; and (iii) patients with BMI ≥27 who have weight-related complications. To assistant healthcare providers with selecting the appropriate medications for their patients, the guidelines also offer a comprehensive table in which all five currently FDA-approved obesity medications are evaluated in terms of their effect on a laundry list of obesity-related complications or other co-morbidities, including prediabetes, type 2 diabetes, cardiovascular disease and risk factors, chronic kidney disease, and mental conditions, to name a few. All five obesity medications are considered “preferred” for use in patients with type 2 diabetes, but Arena’s Belviq (lorcaserin) and Orexigen’s Contrave (naltrexone/bupropion extended-release) are designated as “use with caution” in patients with prediabetes due to “insufficient data for type 2 diabetes prevention.” Vivus’ Qsymia has much better prevention data. AACE/ACE also included tables with comprehensive information on the year of FDA approval, mechanism of action, clinical trials, dose, common side effects, contraindications/safety concerns, monitoring, and comments for each currently approved obesity medication. We are pleased to see these data in what seems to be a fairly easy-to-use, digestible format – we’ll look forward to seeing HCP feedback. As obesity is such a complex disease to treat that requires individualized care, we believe that these data will help facilitate clinical decisions for healthcare providers treating people with obesity.
  
  • Both Drs. Jeffrey Mechanick (Mt. Sinai Hospital, New York, NY) and Timothy Garvey emphasized that the choice of pharmacotherapy should be patient-centered and the success of the therapy depends on patient buy-in. Dr. Mechanick drove home the point that patient input should be guiding all decision-making in obesity treatment as experience with obesity and other chronic diseases demonstrates that physicians simply telling patients what to do doesn’t tend to work. That said, of course, we would want to make sure that the patient choices were guided by education. In addition, he emphasized that all of the pharmacotherapies need to be available in the provider’s toolbox in order to provide optimal, individualized care for patients as currently, healthcare providers and patients are suffering due to lack of access. In particular, he underscored the need for reimbursement for all of the obesity pharmacotherapies across all health plans – that is a naïve hope in our view and we do not anticipate this happening. Reimbursement has been an ongoing barrier for obesity pharmacotherapies at all – as Dr. Dave Dixon (Virginia Commonwealth University, Richmond, VA) shared at ACC 2016, he sees some patients using Novo Nordisk’s anti-diabetes medication Victoza (liraglutide 1.8 mg) for weight loss instead of its anti-obesity sibling Saxenda (liraglutide 3.0 mg) because it is more readily available to some – although presumably, anyone with pre-diabetes instead of diabetes would not be able to access Victoza even though it is the same compound but cheaper by the unit as a diabetes therapy.
• The guidelines offer suggested weight loss and treatment goals based on different weight-related complications of obesity, including prediabetes and type 2 diabetes. For prediabetes and metabolic syndrome, the guidelines recommend a 10% weight loss goal to prevent type 2 diabetes. For patients with existing type 2 diabetes, the guidelines recommend a 5% to ≥15% weight loss with clinical goals of reduction in A1c, reduction in number/doses of diabetes medications, and/or diabetes remission in patients with short diabetes duration. These guidelines help solidify weight loss as an approach to improve patient health and to prevent progression to related comorbidities.
• The very-comprehensive treatment algorithm also includes detailed evidence-based guidance on the anthropometric and clinical components of diagnosis, a detailed checklist of weight-related complications, and specific guidance on potential components of lifestyle therapy. Within lifestyle therapy, the guidelines offer specific strategies for meal plans, physical activity, and behavior modification and also list non-physician healthcare professionals who may be able to participate in creating a team-based model of care of this front – these include dietitians, health educators, physical therapists, exercise trainers, and clinical psychologists or psychiatrists. The emphasis on behavioral intervention and team-based care was particularly impressive to us as lifestyle therapy can sometimes focus on diet and physical activity rather than the psychosocial aspects of obesity.
• Dr. Garvey emphasized the AACE obesity care model that includes a variety of systemic factors that are necessary for long-term, sustained obesity management from a population perspective. He stressed the need for a healthy built environment, a reformed healthcare system, an activated patient who is supported by a prepared obesity practice, improved population based outcomes, and future innovations to improve the state of obesity. He summarized the need for a society solution to work toward a chronic care model for obesity.
• The obesity guidelines and algorithm echo many of the principles of the complications-centric model for the care of patients with overweight or obesity that were summarized in brief in the AACE 2016 diabetes treatment guidelines. The diabetes guidelines similarly suggested a complications-based model to evaluating the need for obesity therapies and suggest a stepwise approach of lifestyle therapy, followed by pharmacotherapy, followed by bariatric surgery as a last resort. We hope that providers will carefully consider both the diabetes and obesity treatment algorithms as complementary guidelines when treating patients with both conditions.

Novel Developments in Diabetes Technology

Panel Discussion

Ed Damiano, PhD (Professor of Biomedical Engineering, Boston University and President and CEO, Beta Bionics, Boston, MA) and Scott Lee, MD (Medical Director, Medtronic Diabetes, Northridge, CA)

Below, we include the transcript from a fantastic Q&A session following presentations by Drs. Damiano and Lee.

Dr. Bob Vigersky (Medtronic Diabetes, Washington, DC): Is the Bionic Pancreas going to take the endocrinologist out of the picture and send patients with type 1 to family physicians and PCPs? What do you see as the role of the endocrinologist once this device is fully operational?

Dr. Damiano: I think the role between providers and patients will change profoundly, and it ought to. I hope this device we are building isn’t too long lived either, because I’m hoping a biological cure will come in my lifetime. I think the focus of the patient encounter for endocrinologist where the bionic pancreas is concerned will be to help people live with this device. Everyone with diabetes can benefit from this technology, but we will have resistance from people who aren’t going to want to wear a device. We will also have to learn how to use this in type 2 diabetes, pregnancy, etc. People who have lived with type 1 diabetes are going to love it, but people with new diagnoses will hate it. It will be hard to explain to every person who is newly diagnosed that this is what he or she is going to have to use every day.

Q: Can you provide a bit more information on the use of glucagon in your system?

Dr. Damiano: In all of our clinical trials to date, we have been using Lilly’s reconstituted glucagon. It’s not chemically stable, so we have an IND exemption that allows us to use this compound and reconstitute it every day. However, there are really nice long-term solutions on the horizon. Xeris and Zealand have stabilized profiles that are as chemically stable as, if not better than, insulin. They’ll remain stable in a pump for half a week or a week.  The nice thing about glucagon is that it’s absorbed very quickly. Reformulations of glucagon are absorbed in the about 15 minutes, so their Tmax is 15 -20 minutes. You’re starting to see an effect in 5-10 minutes; it’s absorbed as fast as orange juice. Glucagon development over past year has made tremendous progress. You’ll hear more about this over the next few months.

Q: One of the most interesting things I took away from a prior talk is the variability in different patients’ absorption peak of insulin. Can you repeat some of those comments?

Dr. Damiano: In the past, I’ve talked about early inpatient studies that have had high data density because we had IVs placed where we could measure plasma glucagon and plasma insulin levels. We found a remarkable 4- to 5-fold intersubject variability in the absorption of Humalog – the Tmax we found was anywhere between 45 and 210 minutes in a small cohort of < 20 subjects. We found that the rate of absorption correlated with in direct proportion with anti-insulin antibody titers.  It also showed a correlation with people who have had type 1 diabetes for many years and had been exposed to the so-called “dirty insulins” – pork and bovine insulins – which had higher immunogenicity than regular human insulin and insulin analogs.  My clinical collaborator, Dr. Steven Russell, has hypothesized that slower absorption in the presence of high anti-insulin antibody levels may have to do with antibodies binding up the insulin analog and sterically hindering the absorption through the subcutaneous tissue.  However, we also found that even in the same person, there was as much as a 50% intra-subject variability. Insulin variability is a huge challenge, but our device uses an average Tmax value that is representative of the T1D population as a whole and manages to do a very good job despite the inherent variability that exists between subjects and within subjects.

Q: From a patient safety standpoint, what have you done in terms of testing in the in-patient setting or in conditions (e.g., lipodystrophy, chronic kidney problems) that could impact the performance of the system?

Dr. Damiano: On the first question, we have built and conducted one clinical trial testing a system that administers insulin and dextrose IV. It uses the same kind of mathematical algorithms as our outpatient system. Just for context, I’d point out that the G5 CGM has acetaminophen interference whereas the Navigator system that we used previously did not have that problem.  We switched over to the G5 when the Navigator was discontinued in the US. In any case, Steven Russell did a study where he compared how the Navigator performed at different frequencies of calibration in critical care environments. He looked at different ICUs and found that if you calibrated the Navigator roughly once every nursing shift, you could get accuracy comparable to that seen in the real world. Essentially, the findings suggested that certain sensors are quite insensitive to interfering substances such as acetaminophen. I understand that the Dexcom technology is moving in this direction as well.  Incidentally, using intravenous insulin and dextrose infusion, our system was able to achieve glycemic control that rivals endogenous control even though we were using transcutaneous CGM sensing rather than venous blood glucose sensing as the input to the controller.

Dr. Lee: Certainly, I think this whole issue is a really good point. A lot of patients have been wearing insulin pumps for 20 years or longer and you’re going to need something in the background that is adjusting for this stuff. As you mentioned, in closed-loop therapy in patients with renal insufficiency, it is going to be important to figure out how to bring blood sugars down and reduce the risk of hypoglycemia. And in insulin-only algorithms, there is an adjustment that can be made to address that.

Dr. Damiano: The one other thing I would add is that Steven Russell is looking at any adverse effect alcohol consumption may have on glucagon performance because that’s something else we need to understand. Do high blood ethanol levels influence glucagon efficacy?

I do also want to address one other idea that has been percolating in the media recently – namely, that insulin-only and bihormonal systems are somehow in tension. They are ABSOLUTELY not in tension. They are complementary. We believe that each system is going to be right for different circumstances and different patients and that certain populations may start on insulin-only, and then may transition to the bihormonal system (people with T1D, for instance).  Other patients may remain on an insulin-only system indefinitely (people with T2D who need exogenous insulin, for example). We need to embrace both systems where there is benefit to be added. In turn, these systems will give way to a biological cure; these are just stepping-stones, after all, to the ultimate solution that we’ve been waiting for. But they are not at odds with one another. We’re seeing temporal solutions that are going to transition to a biological cure.

Q: I’m a heavy user of Dexcom, and from a practical point of view, the sensors aren’t that easy to deal with and aren’t that accurate. My blood glucose can be 150 points different from what my Dexcom says. I also have days where I have three question marks and no reading. If sensors are key to your devices working, then they won’t be very functional.

Dr. Damiano: First of all, there’s no doubt at all that sensors aren’t perfect. Fingerstick meters are also misleading, especially with pediatric patients who often have sugar on their fingers.  With fingerstick meters, you are much more invested in each fingerstick measurement than with CGM – you put all your eggs in one basket for BGM, and then bolus based on that, which is scary. One of the most popular BGMs is much less accurate than the Dexcom CGM. These devices will get more accurate (through improved sensor technologies, new sensor modalities, and sensor redundancy), but people are in desperate need of better technologies right now. We shouldn’t hold this technology back just because it’s not perfect yet. It will continue to improve, but it will never be perfect. We have to acknowledge that we are dealing with an at-risk population and that automated glycemic control systems and automated insulin delivery systems are producing far better control than the current standard of care.

Dr. Lee: One of the things we did was induce an over-calibration of the sensor, meaning that we falsely caused the sensor to read abnormally high glucose levels. If the PID algorithm could correct this unconstrained, then it could obviously lead to low blood glucose. It’s a balance of constraining the system in case of over-calibrating, but also bringing down blood glucose. We decided that it’s important to have a sensor integrity check, and the Medtronic system will have multiple sensor integrity checks. These are all critical issues and it’s important to stress the system and make sure we develop a safe closed loop system.

Dr. Damiano: Sensor redundancy is something people have been thinking about for years. Things will get better, but I never would suggest that these devices won’t overdose people and give them severe hypoglycemia. It will happen in outlying situations. But what is going on right now with current standard-of-care therapies is terrible. The standard of care is so easy to beat because it is, unfortunately, so bad.

Q: Are you thinking about using the Bionic Pancreas in populations other than type 1 diabetes?

Dr. Damiano: We do see tremendous opportunities for patients outside of the type 1 diabetes community – people with type 2 diabetes, pancreatic cancer patients, patients with congenital hyperinsulinism, patients with post-bariatric hypoglycemia. We haven’t tested the device in populations outside of type 1 yet (except for some preliminary experiments in type 2 diabetes), but we expect that it will perform well in these populations.

Q: How often do you have to calibrate the user’s weight?

Dr. Damiano: For children that expect to gain significant amounts of weight, the Bionic Pancreas will give you the option to have it remind to update your weight after a period of time (say 6 to 12 months). However, in an adult whose weight typically changes only 20-30 pounds, changes in weight are not going to make much of a difference. After all, when you look at the results of our trials to date, you’ll see that our device has figured out how to adapt insulin dosing for everyone from pre-adolescents to adolescents, to adults in about 18 hours, regardless of weight differences.  In particular, we have seen our system adapt in two subjects who weigh the same but have different insulin requirements.  Nevertheless, the system was able to automatically adapt insulin dosing appropriate to the needs of both individuals.. Ultimately, you’ll typically need to update weight only in the case of young children as they go through their growth spurts. 

-- by Helen Gao, Varun Iyengar, Sarah Odeh, Emily Regier, Ava Runge, and Kelly Close