In diabetes therapy highlights, Dr. Alice Cheng shared her personal flowchart for aiding prescribing practices in light of the torrent of recent CVOT data that has emerged in recent years, Drs. Stephen Hull and Peter Senior faced off in a debate on metformin’s place as a first-line therapy, and Dr. Irene Hramiak provided an encompassing overview of the history and future of insulin therapy. We also heard from Dr. Lawrence Leiter on the complex relationships between the triad of glycemic variability, severe hypoglycemia, and CV outcomes. Dr. Sean Wharton also provided an engaging keynote, providing several calls to action to confront the incoming “obesity tsunami” that we will soon face.
In diabetes technology highlights, Dr. Richard Bergenstal provided a powerhouse on the CGM-enabled transformation in diabetes care, Dr. Medha Munshi spoke about the unique challenges faced in the management of elderly type 1s, and Dr. David Kerr discussed the usual digital health buzz topics but in the context of active programs running out of Sansum.
The Diabetes Canada exhibit hall was also brimming with activity, due to Health Canada’s approval of Medtronic’s MiniMed 670G and Tandem’s t:slim X2 with G5 integration. On the therapy front, we gained some valuable insights into when full results for CREDENCE may be released, Tresiba coverage extending in certain Canadian provinces, and recent rollouts of Ozempic and Soliqua in Canada.
Below, you’ll find all of our coverage of a learning-rich Diabetes Canada 2018 meeting in Halifax, Nova Scotia this October. Our top daily highlights are divided into diabtes therapy, diabetes technology, and obesity/psychosocial/big picture highlights. Read on!
- Diabetes Therapy Highlights
- Dr. Alice Cheng Provides Personal Flowchart for Practically Integrating CVOT Data into Clinical Practice; She’ll Retire if CAROLINA Reports More CV Events in SU Arm Than DPP-4 Arm
- Titans Debate: Should Metformin Remain First-Line Therapy in Type 2 diabetes? Yes, But with Some Qualifications.
- Dr. Irene Hramiak Says Only 25% of Canadian Family Physicians Prescribe Insulin, Pushes for CDE Insulin Prescribing; Sweeping Overview of Past, Present, Future of Insulin
- The Glucose Variability-Hypo-CV Triad: Dr. Leiter Reviews Literature on Associations
- Drs. Jordan Weinstein and Alice Cheng on Kidneys and Diabetes: CREDENCE’s Early Halt First Exciting Nephrology News in 17 Years
- An Endocrinologist, Diabetes Nurse Educator, and Psychiatrist Lay Out Tips and Tricks for Insulin Initiation
- SPRINT Trial – Follow-up to ACCORD-MIND – Demonstrates Benefits of Intensive Blood Pressure Lowering in Preserving Cognitive Function; Could There be NOTs (Neurologic Outcomes Trials) for Diabetes Agents?
- Diabetes Technology Highlights
- Dr. Bergenstal’s Hit Keynote Featuring Beyond A1c, CGM, and New Tech Talking Points; Doesn't Ask for A1c at Appointments Anymore!
- Sansum’s Dr. David Kerr Overviews Digital Health-Enabled Efforts to Address All Five Determinants of Health; Big Data (Worse Sleep, Fewer Steps) Predicts Flu; Diabetes Needs to Focus More on Time as Currency
- Joslin’s Dr. Munshi: Considerations for Caring for the Elderly Type 1; “Not the Best Way to Treat Diabetes by Far, but it is the Best Way to Treat YOU”
- “Not Another Boring BGM Talk” Delves into BGM/CGM Accuracy, Heightened SMBG Frequency with FreeStyle Libre?
- Drs. Sherr & Aleppo Give Positive Commentary and Tips at Medtronic Symposium Two Days Post-Health Canada 670G Approval; First Units to Ship in December
- Obesity, Psychosocial, and Big Picture Highlights
- “Overcome Your Bias” Toward Obesity Pharmacotherapy; Only 2% of People with Obesity Treated with Medication!
- “If There’s Diabetes Look for Depression, and If There’s Depression Look for Diabetes”: Discussion of Diabetes Canada Guidelines Focus on Mental Health Effects; Strong Agreement with ADA/EASD Consensus Guidelines
- Tidbits on Exercising with T1D From Co-Author of 2016 ADA Recommendations; What About Women Exercising with T1D?
- Titans Debate #2: Audience Seemingly Swayed in Direction that Intermittent Fasting is a Simple and Healthy Addition to Management of T2D
- Country Star George Canyon Says Care Network (Including Parents) Should be Corporation, Where Teen with Diabetes is the CEO; Should We Change Nomenclature to Prevent T1D vs. T2D Confusion?
- Young Type 1s Who Experience Stigma 2x More Likely to Have A1c >9% or Severe Hypoglycemia in the Past Year
- Dr. Michael Vallis: Technology Can Bridge the Gap for Sustained Behavior Change
- Exhibit Hall
Diabetes Therapy Highlights
Dr. Alice Cheng (University of Toronto) shared her personal flowchart for aiding prescribing practices in light of the torrent of CVOT data that has emerged in recent years. We believe the below chart should be distributed far and wide, particularly to GPs, and kept up to date: It helps providers decide which cardioprotective diabetes agent to use and when (she did not consider albiglutide, considering GSK recently pulled it from market) in patients with type 2 diabetes and established cardiovascular disease. Dr. Cheng first considers kidney function: if patients have an eGFR rate < 15 mL/min/1.73 m2, then she does not add any of these therapies, seeing as none of them have been indicated in those with severe renal impairment. If the patient has an eGFR between 15-30 mL/min/1.73 m2, then either of the two GLP-1 agonists are used (liraglutide or semaglutide), as both EMPA-REG and CANVAS for the SGLT-2 inhibitors excluded patients with an eGFR below 30 mL/min/1.73 m2. At an eGFR greater than 30 mL/min/1.73 m2, any of the four agents can be used; therefore, Dr. Cheng then looks to see if the patient has established or high risk of heart failure. If so, either of the SGLT-2 inhibitors should be used, since clear HF risk reduction has been seen with their use in EMPA-REG and CANVAS. Otherwise, any of the four agents can be used – we imagine a combination of reimbursement, cost, patient preference for a pill vs. injectable, and provider preference will drive this decision. Dr. Cheng then proceeded to the bottom half of her flowchart by providing the caveat that this section waded into an area filled with her opinions, rather than cold-hard facts. After one of these agents has been chosen, Dr. Cheng likes to consider whether their addition will bring about increased hypoglycemia risk for the patient. To do this, she sees whether the patient is currently on a sulfonylurea or insulin (which both harbor considerable hypoglycemia risk); if they are not, then no adjustment is needed, and treatment can proceed by simply adding the decided-upon agent. If they are, then Dr. Cheng considers the A1c of the patient: if it’s ≤ 8%, then she reduces the sulfonylurea dosage by 50% or stops it altogether (no advice was given on what to do with insulin titration); otherwise, no adjustments are needed. To be clear, these recommendations are surely not approved guidelines, but we do find them quite valuable nonetheless. We’re glad to see a thought leader like Dr. Cheng comprehensively incorporate the ins-and-outs of CVOT data in such a practical and easy to follow manner into their practice. We also greatly appreciate her consideration of hypoglycemia risk with the addition of these agents in certain patients.
- In a light-hearted moment when discussing upcoming results from the CAROLINA trial assessing CV outcomes between linagliptin and the sulfonylurea glimepiride, Dr. Cheng went on the record and said that “if CAROLINA shows that sulfonylureas have fewer CV events than DPP-4 inhibitors, then I will retire, because nothing will make sense at all.” We too would be shocked with this result—so don’t expect to see Dr. Cheng hang up her white coat any time soon ;>!
- Starting off her talk by confessing that she would be providing a “rose colored picture” of CVOTs that focused on the many benefits that have emerged from these trials, Dr. Cheng delivered a thorough overview of the studies to date. She began with the CVOTs that have demonstrated CV safety of DPP-4 inhibitors (EXAMINE, SAVOR-TIMI, TECOS, and as of last week, CARMELINA), and then moved into the more exciting realm of positive CVOTs in the SGLT-2 and GLP-1 classes. Dr. Cheng underscored that to date, five CVOTs have shown positive CV effects: EMPA-REG for empagliflozin, CANVAS for canagliflozin, LEADER for liraglutide, SUSTAIN-6 for semaglutide, and HARMONY-Outcomes for albiglutide. At this point, these trials have been discussed at length (although they never seem to run out of providing incredible insights!).
An entertaining debate between Dalhousie University’s Dr. Stephen Hull and University of Alberta’s Dr. Peter Senior on the use of metformin as a first-line therapy in type 2 diabetes failed to sway audience members from their baseline assessments: In both the pre- and post- poll, ~70% believed that metformin should still be used as a first-line therapy. Both men got some good punches into their preliminary arguments and rebuttals. However, the debate really ended up meeting somewhere in the middle, and appropriately so. Tasked with arguing in favor of metformin as the first-line therapy, Dr. Hull was in a tough spot; Dr. Senior had the luxury of advocating for different first-line medications, combination therapies, and a more personalized approach. His conclusion stated that “metformin should remain first line therapy when combined with other agents as part of an individualized approach, agreed with the PWD, to achieve their long-term health goals.” Dr. Hull didn’t have the chance to respond to this point, but we’re almost certain he would’ve agreed.
- Dr. Hull argued the pro side first, plainly stating that he’s “still in love with metformin.” His first point was that all guidelines suggest that metformin should be first line (a moot argument, in our view, since the debate essentially boiled down to whether the guidelines are right or wrong). He then turned to the literature: In the UKPDS, the small number of type 2s with overweight who took metformin had a lower risk of adverse events; in Spread-DIMCAD, there was an adjusted HR of 0.54 for MACE (+revascularization) for individuals taking metformin vs. glipizide; in HOME, adding metformin to insulin significantly reduced macrovascular events (HR: 0.61); and in the Reach registry (n~20,000), all-cause mortality was 6.3% for those on metformin and 9.8% for those not on metformin. He laid out a number of the recognized advantages of metformin, including that it’s been in clinical use for >60 years (the safety profile is well-understood), has high A1c-lowering, low cost (if everyone in the US was on a novel diabetes agent, it would cost $139-$270 billion per year), negligible risk of hypoglycemia, weight neutrality, and a mild side effect profile. Beyond hepatic mechanisms, Dr. Hull pointed out potential cancer prevention/aging benefits, effects on the gut (hormones, gut/brain axis, microbiome), and benefit on CV risk factors and atherothrombotic factors.
- Dr. Senior’s pitch came down to his feeling that metformin’s place in guidelines may be a stubbornly persistent quirk of history, and a simplified treatment algorithm for a complex disease. Firstly, he noted that there are other medications that have low risk of hypoglycemia, no weight gain (plus weight loss!), and evidence of CV safety (plus CV protection) – many of metformin’s’ ticked boxes, minus low cost in the branded cases. [Dr. Hull later pointed out that new drugs haven’t been studied as initial monotherapy, and >70% of people enrolled in many major CVOTs for novel agents were already on metformin as a first-line]. Secondly, diabetes is a very heterogeneous condition, so no one therapy should be selected or all patients, said Dr. Senior. Third, and most convincingly, there is actually no strong evidence for the supremacy of metformin as a first-line therapy in type 2 diabetes. Metformin-first guidelines have deep roots in the UKPDS, where people with overweight were randomized to an intensive arm (insulin or sulfonylurea or metformin) or conventional. Dr. Senior characterized the conventional arm as “standard neglect,” as the aim was simply to keep fasting plasma glucose <270 mg/dl. Continuing his dismantling of UKPDS’ metformin findings, Dr. Senior pointed to a number of methodological problems with the trial, including unblinding of endpoints, addition of sub-studies, extensions of trials, and extrapolation from sub-analyses. The final shot was that a sub-analysis shows adding metformin to sulfonylurea therapy is actually worse for all-cause mortality and diabetes-related deaths than sulfonylurea alone. Lastly, the UKPDS authors wrote that metformin could be chosen as a first line therapy, and cautioned that their findings might not apply to non-overweight groups. Thus, from the UKPDS, Dr. Senior can only conclude that metformin should remain first line if compared to conventional therapy (“standard neglect”) and if the only other choices are SUs and insulin. Dr. Senior’s final point was that diabetes is a lifelong disaease, and monotherapy is short-sighted. Invoking UKPDS once more, he pointed out that “for 10 years, metformin does nothing (for CV events). That’s the time you’ll be entering a CVOT. If the drugs are so powerful, why wait 10 years to use them?” “Metformin should remain first line therapy if you only care for patients for five years,” he said, citing the ADOPT trial where A1c rapidly dropped in the metformin group but returned to baseline five years later. “Take Mr. 40 – 40-year-old, BMI of 40 – do you really think that metformin today vs. something else or a combination will really shift the needle to increase life expectancy?
St. Joseph’s Dr. Irene Hramiak delivered the first plenary lecture of Diabetes Canada 2018, reflecting on the history of insulin R&D and identifying existing care gaps. She provided a whirlwind historical tour of diabetes research and care, beginning with early forms of treatment (including prescribed horseback-riding thousands of years ago in order to reduce urination frequency!) all the way to next-generation smart insulins in development today. Despite these monumental leaps, Dr. Hramiak emphasized that inertia in the prescribing of insulins needs to improve above all. She noted that physicians fear hypoglycemia just as much as patients do, and this may contribute to the fact that only 25% of family physicians in Canada who look after people with diabetes prescribe insulin. That figure is truly remarkable, and we wonder (i) how many people guidelines would say should be on insulin and (ii) how it stacks up to HCPs in other countries. Interestingly, Dr. Hramiak advocated for the diabetes educators scope of practice to be expanded to include prescribing insulin as a way to combat inertia under-prescription. This is a brilliant idea, as educators are the ones on the ground who understand patients’ hesitation the best and can guide them through to acceptance and readiness. In terms of modern insulin therapies, Dr. Hramiak identified the central struggle facing HCPs is to balance the competing forces of lowering A1c levels to reduce complications while also trying to mitigate hypoglycemia risks. On this front, newer basal insulins have surely helped: Dr. Hramiak detailed how these novel agents – namely U300 glargine (Sanofi’s Toujeo) and insulin degludec (Novo Nordisk’s Tresiba) – have risen to the task. However, ample room remains for improvement, especially in terms of improving flexibility with these agents. Dr. Hramiak pointed out that the lack of flexibility with longer acting basal insulins must be addressed in order to accommodate commonplace events in the patient experience, such as increases in exercise or changes in diets. In her eyes, the ideal basal insulin would show glycemic-lowering efficacy, predictability in dosing, simplicity of delivery in terms of device, frequency, and route, and also harbor a low risk of hypoglycemia.
- Dr. Hramiak spoke positively on the potential of fixed-ratio combination therapies (basal insulin + GLP-1s) early in treatment intensification. When asked by Dr. Alice Cheng (University of Toronto) on where these therapies should fit into treatment algorithms, Dr. Hramiak responded that they could be used “right after metformin.” To be sure, Dr. Hramiak is not exactly sharing a controversial opinion here –she joins a long line of KOLs that have expressed enthusiasm over the effectiveness of these agents. Nevertheless, a deep divide remains between this enthusiasm from thought leaders and the actual uptake of these therapies, namely due to barriers in terms of perception, reimbursement, and labeling rigidity. We’re surely glad to see leaders like Dr. Hramiak support the use of these combination therapies early on in treatment and couldn’t agree more with her sentiment.
- This enthusiasm for fixed-ratio combination therapies was echoed later in the day by Dr. Stewart Harris (Western University, London, Ontario). Notably, these agents (Xultophy; Soliqua) have just emerged onto Canadian markets in the past few months following approval earlier in 2018. Dr. Harris commented that he expects to see the decline of bolus therapy in favor of these combinations, and highlighted that these combination therapies are “the most potent and effective injectables that we have on the market” – wow! Dr. Harris also positively remarked that in his practice, more and more patients are being taken off of basal/bolus therapies in favor of combination therapies – we view this as such a positive sign and hope to see more and more patients reaping the tremendous benefits of these therapies.
The Glucose Variability-Hypo-CV Triad: Dr. Leiter Reviews Literature on Associations
In an overflowing evening symposium, University of Toronto’s Dr. Lawrence Leiter reviewed the literature on the relationships between glycemic variability, hypoglycemia, and CV risk/mortality. At the end of the lecture, pre- and post-polls showed that attendees were moved to buy that the three are correlated, but causality still remains a point of debate, especially for the effect of variability or severe hypoglycemia on CV risk. Since ethical concerns would seem to obviate an RCT where people are randomized to have more severe hypoglycemia or greater glycemic variability, the link may never be fully understood, though the association alone suggests that variability and hypoglycemia should both be addressed. Below, we delve into the three sides of the triad, as elucidated by Dr. Leiter.
The relationship between severe hypoglycemia and risk for MACE/all-cause mortality.
Severe hypoglycemia has been shown to be associated with MACE and all-cause mortality across CVOTS: VADT, ACCORD, ADVANCE, EXAMINE, ORIGIN, LEADER, and DEVOTE. Examples: (i) In ACCORD, patients in the intensive arm who had a history of severe hypoglycemia had a ~3x greater all-cause mortality; (ii) In ORIGIN, patients with history of severe hypoglycemia had a 1.6-1.8x risk of adverse CV outcomes; (iii) In DEVOTE, patients with a history of severe hypoglycemia were ~2x more likely to die of CV disease, and ~2.5x more likely to die period. A CVOT meta-analysis by Yeh et al. made the case for a link between hypoglycemia and adverse outcomes (all-cause mortality, CV mortality, and a major CV event). BUT…there is compelling evidence that severe hypoglycemia may be simply a marker of frailty: In ADVANCE, patients with a history of severe hypoglycemia were more likely to have micro- and macrovascular events, but they were also more likely to have respiratory, digestive, and skin disorders. Said Dr. Leiter, “It’s hard to imagine that severe hypoglycemia causes patients to have skin disorders, so this analysis suggests it may be more vulnerable, sicker patients having severe hypoglycemia.” While in LEADER and DEVOTE, patients were more likely to die in close proximity to a severe hypoglycemia event, which can be taken as evidence for hypoglycemia being both a marker or a mediator of an adverse outcome.
- A debate at IDF 2017 between Drs. Brian Frier and Prof. Markolf Hanefeld concerned the very topic of causation vs. correlation. Though Prof. Frier argued in favor of causation, but suggested that an expensive, labor-intensive prospective study using simultaneous CGM and Holter monitoring (EKG; maybe Apple Watch Series 4 now?) would be the only way to firmly establish causation. (We’d note Dr. Irl Hirsch tried this in the FLAT-SUGAR pilot study, but showing big differences in glycemic variability/hypoglycemia for the same A1c was challenging.) Prof. Frier believes a trial would actually reveal that the answer to the debate probably lies “somewhere in the center ground.”
The relationship between glycemic variability and hypoglycemia.
This is the most stable side of the triangle, as high glycemic variability naturally correlates with more hypoglycemia. In a Lancet D&E paperpublished in August, Ceriello et al. showed that both type 1s and type 2s with more glycemic variability had more hypoglycemia. This relationship held for low mean glucose (<150 mg/dl), medium mean glucose (150-180 mg/dl), and high mean glucose (>180 mg/dl). Drs. Kovatchev and Cobelli showed similar data in a 2016 Diabetes Care paper. When it comes to large RCTs, more variability correlated with more frequent overall, nocturnal, and severe hypoglycemia in SWITCH, and more glycemic variability correlated with more severe hypoglycemia in DEVOTE.
The relationship between glycemic variability and MACE/all-cause mortality.
There is less data here, though some emerging. Diabetes Care recently published a paper from Zhou et al. showing that those in the intensive arm of the VADT with more glycemic variability were more likely to have an adverse event. In DEVOTE, those with greater day-to-day fasting glycemic variability were more likely to have severe hypoglycemia, MACE, and all-cause mortality. Thus, the association in type 2 seems robust. However, in type 1, a paper published in 2017 (Lachin et al.) found that glycemic variability had no relationship to retinopathy. Of course, there could be significant undetected glycemic variability in DCCT that a CGM would’ve picked up.
Nephrologist Dr. Jordan Weinstein (St. Michael’s Hospital) and Dr. Alice Cheng (University of Toronto) offered unique insights on kidney care in diabetes management during a Janssen-sponsored breakfast symposium. This early-morning symposium was rife with insights on the kidney in diabetes care. See below for our favorite learnings!
- That CREDENCE was stopped early for significant renal protection represents the first truly exciting news for nephrologists in 17 years. Dr. Weinstein noted that until earlier this year, his presentation of this symposium on renal function in diabetes would have essentially been the exact same for the past 17 years. Seventeen years ago was the last time that a medicine was approved with an indication for kidney disease for patients with diabetes (when both Irbesartan and losartan were approved), and Dr. Weinstein remarked that he has been waiting for exciting news on this front ever since. In this context, CREDENCE has certainly delivered. As a reminder, the study of J&J’s SGLT-2 inhibitor Invokana (canagliflozin) was terminated early for the positive reason of meeting its primary endpoint in demonstrating renal protection in type 2 patients with DKD. This announcementwas released in July 2018, and we expect to hear full results from the study sometime in 1Q19. Dr. Weinstein remarked that CREDENCE results are “like Christmas – you know you’re going to get something good, and now you’re just waiting for it.” Dr. Weinstein also noted that the simple design of CREDENCE will allow for its seamless integration into practice. We definitely cannot overstate the importance of and excitement around CREDENCE, and are glad to see thought leaders share in this enthusiasm—it has been so long since major progress has been made on this front!
- CREDENCE will build on the story hinted at by secondary endpoints in EMPA-REG and CANVAS that demonstrated renal risk reduction. Both of these CVOTs contained multiple renal endpoints as part of their secondary analysis (EMPA-REG: composite of doubling of serum creatinine, initiation of renal replacement therapy, or renal death; CANVAS: composite of doubling of serum creatinine, ESRD, or renal death) and both showed a significant risk reduction on these endpoints (46% for EMPA-REG and 41% for CANVAS). Needless to say, as secondary endpoints, these results reflected a small number of total events, necessitating the need for the larger CREDENCE study. Nevertheless, the totality of these results does point favorably towards a positive class effect of SGLT-2 inhibitors with renal protection. Both empagliflozin and dapagliflozin will be more thoroughly studied in EMPA-KIDNEY (to be started shortly in November 2018) and Dapa-CKD (began February 2017) in this regard; results from these larger trials will help to definitively answer questions about the potential class wide renal protective effects of SGLT-2 inhibitors.
- Unlike other diabetes complications, little progress has been made in preventing end stage renal disease (ESRD). Rates of complications such as stroke, acute myocardial infarction, and amputation have all seen significant declines since 1990; however, ESRD rates have showed only a slight decline (see graph below). Moreover, 40%-50% of people with diabetes will develop diabetic kidney disease (DKD), which is the leading cause of ESRD. Even more shockingly, between 1990 and 2012, the number of deaths due to DKD rose by 94%. While a significant chunk of this phenomenon is certainly attributable to improved healthcare leading to increased longevity of patients and the eventual emergence of renal dysfunction as a result, Dr. Weinstein still highlighted how much of a burgeoning problem this represents, and the enormous unmet need for therapies to address this issue.
- On the distinction between CKD and DKD: Both Drs. Weinstein and Cheng joked about the proliferation of acronyms in the diabetes landscape, and how overwhelming the spread of jumbled letters can often appear. To clear the air, Dr. Weinstein provided an extremely helpful explanation on the difference between these two often interchanged terms (chronic kidney disease [CKD] vs. diabetic kidney disease [DKD]). He explained that DKD encompasses both glomerular conditions (i.e. “diabetic nephropathy” that is marked by proteinuria) and atherosclerotic conditions (i.e. “ischemic nephropathy” marked by inflammation and lack of proteinuria) that affect the kidneys as a result of diabetes. CKD is a broader term and includes DKD, but also brings in other conditions that affect the kidneys, such as hypertensive nephrosclerosis, glomerulonephritis, interstitial nephritis, and multiple myeloma.
- The distinction between the CKD and DKD is not merely a trivial construction. Instead, Dr. Weinstein also explained how realizing the difference between CKD and DKD can inform patient care: “if you realize you’re looking after a patient with atherosclerotic DKD, you can assure them that they won’t progress as quickly [in renal dysfunction] – the lack of proteinuria is definitely encouraging.”
In a Lilly/BI-sponsored symposium, renowned endocrinologist Dr. Alice Cheng, diabetes nurse educator Trish Collins, and psychiatrist Dr. Barry Simon provided strategies for initiating patients on insulin therapy. Each of these impressive individuals took the stage to share insights from their experiences on how best to tackle the complex problem of starting patients on insulin—an extremely topical discussion, especially given Dr. Irene Hramiak’s statistic this morning (see above) that only 25% of family physicians in Canada who see people with diabetes prescribe insulin.
- Dr. Barry Simon (University of Toronto) provided tips and tricks on how to deliver holistic care that places an emphasis on mental healthcare. When discussing insulin initiation, caregivers must provide E.M.P.A.T.H.Y., also addressing underlying eating disorders, mood disorders, pain disorders, anxiety disorders, trouble sleeping, hyperactivity and attention issues, and your alliance with the patient. We also found Dr. Simon’s approach to shame attenuation to be helpful: In conversation with patients, he recommends using the language of “many people” to minimize shame that the patient may experience about starting insulin therapy. For example, an HCP can discuss the patient’s thoughts about the therapy by framing the question as, “Many people have concerns about starting insulin. What are yours?” Such a strategy helps the patient feel as a part of the majority, rather than isolated and shame-ridden.
- The great Dr. Alice Cheng (University of Toronto) reviewed actions that clinicians can take to ease patients into insulin therapy. She leveraged results from the EMOTION study, which aimed to identify specific HCP statements and actions that helped patients overcome psychological insulin resistance. In the study, 594 patients who initially displayed significant psychological insulin resistance were given an online survey evaluating the occurrence and degree of helpfulness of HCP statements and actions in the context of insulin therapy initiation. The results were clear: patients overwhelmingly saw certain actions as very helpful (such as the HCP demonstrating the insulin injection process during their meeting, explaining the benefits of insulin, and adopting a collaborative style with the patient), while finding authoritarian communication styles from their HCPs to be not helpful at all. On this note, Dr. Cheng offered five “practical considerations” to help in discussing insulin initiation with patients: (i) address known perceptions of insulin therapy, (ii) offer an “out” for the patient – do not lock them down and give them no choice!, (iii) keep the regimen simple and negotiate with the patient, (iv) show the patient the needle and device, and (v) provide a frame of reference for dosing (i.e. tell the patient that although we’re starting at 10 units, this is a tiny amount and don’t be afraid to increase).
- Finally, diabetes nurse educator Ms. Trish Collins ran through a litany of techniques she uses in her appointments where insulin initiation is discussed. What we found most salient was her suggestion to demonstrate the insulin injection in the office, and to do this as early as possible in the appointment with the patient. She emphasized that patients have such a great fear of the injection that they don’t absorb what’s being said in the meeting until after they have experienced the feeling of the insulin pen and the injection themselves. After this has been done, they show a diminished sense of fear and are able to absorb the important information being presented.
Highlighting a “Diabetes and the Brain” symposium, Wake Forest’s Dr. Jeff Williamson provided a sneak peek of soon-to-be-published data from the SPRINT trial, which had secondary endpoints investigating the potential of intensive blood pressure lowering treatment to reduce incidence of dementia and mild cognitive impairment (in people without diabetes). Dr. Williamson was previously involved with the ACCORD-MIND trial, a subset of the ACCORD trial showing that aggressive glucose and SBP (systolic blood pressure) control did not slow memory decline. However, ACCORD-MIND did provide some hints at a potential link between intensive blood pressure control and improved cognitive function, as white matter disease development (a marker for dementia) was significantly reduced in the intensive treatment group (p=0.01). Building off of this finding, the SPRINT trial enrolled 9,361 patients and randomized them to standard (SBP <140 mmHg) and intensive (SBP <120 mmHg) blood pressure control. Although the primary endpoints for this trial involved CV events, secondary endpoints investigated adjudicated occurrence of all-cause probable dementia, mild cognitive impairment, and a composite outcome of these two measures. Prompting considerable excitement from the room, Dr. Williamson revealed currently unpublished results from the study (and politely asked for no pictures to be taken, hence the lack of a Kaplan-Meier curve below!). After five years of follow-up, the number of adjudicated occurrences of dementia was lower in the intensive treatment arm (175 events vs. 147 events), but not significant even at 2.6 years of follow-up. However, very positively, the number of adjudicated occurrences of mild cognitive impairment (MCI) was highly significantly reduced in the intensive treatment group (348 events vs. 285 events). This MCI result was also The final secondary endpoint of the combination of occurrences of dementia and mild cognitive impairment was also significant, showing a 15% reduction in both of these events (463 events vs. 398 events). Given the MCI findings, perhaps allowing more time to elapse would show a greater difference in incidence of full-blown dementia. We’ll be sure to keep our eyes peeled as full published results for this trial become available.
- Importantly, no patients with diabetes were included as participants in SPRINT, as diabetes was a specific exclusion criterion. Nevertheless, Dr. Williamson expressed optimism about the potential to apply these results to diabetes care and insisted that important insights about diabetes can be gleaned from this study. To be sure, the link between declining cognitive function and diabetes has attracted considerable interest recently (e.g., Prof. Jaakko Tuomilehto, Dr. Alan Jacobson), and Dr. Williamson also highlighted this link in his remarks (we found most notable his discussion of this 2008 paper detailing a strong correlation between diabetes duration and cognitive processing speed, controlled for external variables). We surely hope that next steps with these findings involve determining whether a similar intensive blood pressure treatment in patients with diabetes would yield similar improvements in cognitive function. Moreover, what role could diabetes medications – especially SGLT-2 inhibitors, which have noted blood pressure-lowering effects – play in this context?
- Dr. Williamson also gave an 0ff-the-cuff remark on the potential for glucose-lowering agents to be tested for effects on brain function – similar to how today these agents must be rigorously tested in CVOTs for CV safety. This idea of a cognitive function trial for glucose-lowering agents is certainly interesting – and provocative – especially in light of recent debates on the necessity of CVOTs and the upcoming FDA Ad Comm meeting to discuss them. While we remain hesitant for such studies to become a mandate, we nonetheless are intrigued by the learnings that could emerge from them. Perhaps, instead, post-market surveillance programs could track medications for signals of neuro-protection and make way for labeling claims. In this new era where diabetes care has strived to move past glucose-centrism (with benefits emerging in terms of time-in-range, weight loss, cardioprotection, and renal protection), could effects on cognitive function represent one of the next frontiers in the beyond-A1c movement? While CV events and death are both terrible and visceral, loss of cognitive function can be just as devastating – if not more so – for an individual and loved ones.
Diabetes Technology Highlights
IDC’s Dr. Rich Bergenstal was true to form in a packed keynote at Diabetes Canada, where he discussed outcomes beyond A1c and technological strides within the framework of six major transformations in diabetes management. The engaging talk, as we’ve come to expect from Dr. Bergenstal, featured significant commentary on the shortcomings of A1c and the efforts to validate CGM metrics – he said that he doesn’t even ask his staff to provide him with his patients’ A1cs anymore! – the new term for eA1c (Glucose Management Indicator or GMI, which he hopes will be introduced into commercial data management platforms in the coming weeks and months), and the quantum leaps in technology (MiniMed 670G was approved by Health Canada yesterday). The six steps that Dr. Bergenstal has observed (along with his supporting evidence and comments) are:
- (i) Fear to Hope. Fitting to the Canadian venue, Dr. Bergenstal gave a tribute to Dr. Frederick Banting, suggesting that the discovery and implementation of insulin took away the fear of diabetes. He played a clip from Adam Brown’s Bright Spots and Landmines audiobook describing the “darkness in diabetes” (negativity, confusion, blame, guilt, fear, etc.) but he stopped the clip after the line, “We can do so much better.” He then showed an overnight CGM trace from an individual who had a hypoglycemic seizure. Despite numerous alarms, the patient didn’t stir, and meanwhile, the pump’s preset basal rate increased to prevent dawn phenomenon, which the literature says occurs ~56% of the time. Dr. Bergenstal pointed to the just-licensed in Canada 670G, evidence that yes, the field can do better.
- (ii) Population health (A1c) to Personalized care (CGM/time-in-ranges). Dr. Bergenstal comically explained how he’s capitalizing on the lack of consensus on A1c targets between AACE (≤6.5%), ADA (≤7%), and ACP (7%-8%) to say, “Perfect, no one can agree on an A1c goal. Let’s get rid of it. Let’s look at actual blood glucoses.” (A1c still has value, he added, for research and as a marker for long-term complications.) He spoke to the “amazing” CGM systems available today (Abbott, Dexcom, Medtronic, Senseonics), and the growing body of evidence correlating time-in-range to complications. For example, he alluded to the recent Diabetes Care paper identifying a significant inverse relationship between time-in-range and the prevalence/severity of retinopathy in hospital-based Chinese adults with type 2 diabetes (n=3,262), and hinted that Dr. Beck et al.’s analysis of SMBG-derived time-in-range vs. microvascular complication burden in the DCCT would be published in Diabetes Care next week. Pertaining to personalizing care, we were also very happy to hear in Q&A Dr. Bergenstal’s impression that US payers are interested in conversations around quality of life far more than they have been in the past.
- Dr. Bergenstal polled the audience: Of the before (red; column B) and after (blue; column A) periods, which is more desirable? Though the patient was disappointed with the 0.4% A1c increase, Dr. Bergenstal remarked that he would actually take the “after” period despite the increased A1c and decreased time-in-range (81%->75%), due to the improvement in hypoglycemia (7%->2%). This is the tradeoff and dialogue enabled by AGP, though the patient’s preference for the “before” period underscores the deeply-rooted fixation on A1c.
- Dr. Bergenstal reminded the audience of rules of thumb for time-in-ranges benchmarks and A1c-<->time-in-range conversions. Based on the 670G US pivotal study, he (and other KOLs) believe 72% of time in 70-180 mg/dl, 3% of time <70 mg/dl, and 25% of time >180 mg/dl is possible for type 1 diabetes. As for translating time-in-range to A1c and vice versa, the general rule for the “average patient” is that 70% in-range correlates with an A1c of 7%, and every 10% delta in time-in-range correlates with a ~0.5% change in A1c (e.g., 75% time-in-range would be ~6.5% A1c). One wonders if these correlations are robust enough for FDA and payers to invoke the transitive property and say that high time-in-range is likely to correlate with low long-term complication incidence.
- (iii) Monitoring (time-in-ranges) to Management (AGP patterns). In other words, the numbers (time-in-range, time in hypoglycemia, glycemic variability, etc.) tell you if you need to act, but they don’t tell you how to act, which the AGP visual does. And at the bottom of the totem pole for clinical management, A1c: “The A1c is hopelessly worthless for day-to-day management. The two variables you need, in my humble opinion, are time-in-ranges and hypoglycemia. You can do everything with those two numbers.” He described the AGP has the ultimate shared decision making tool; when a clinician puts an AGP in front of a patient and asks them why they think the trace looks the way it does, “they will tell you instantly. They know why it happened, it’s just that no one’s ever shown them a picture. They’ve just gotten 8.7%. Oh yeah, that’s very helpful.”
- (iv) CGM with pump to CGM with MDI. The whole message here was that CGM is becoming the standard of care (especially in kids <6 years old, where T1D Exchange Registry penetration is 49% as of 2/13/18), and people on MDI also derive benefit. “If you’re in a clinic and decide the patient is ready for technology, what do you reach for: a pump or CGM? I’d say maybe CGM would be a good choice. You learn about your data, learn your patterns, then see if you’re ready to go to the next step, be it a pump or hybrid closed loop.” In support of the CGM-first mindset, Dr. Bergenstal pointed to the DIaMonD study, which showed a 1.0% absolute A1c reduction (-0.6% relative reduction vs. SMBG) at 26 weeks, plus 76 minutes more per day in range, despite the fact that investigators simply handed participants CGM with minimal education – they quickly learned which foods, exercises, and doses worked for which situations on their own. We wonder how much greater the glycemic improvements could’ve been had they received more intensive training, ongoing support, or even decision support! Dr. Bergenstal also implored providers to give CGM to everyone with impaired awareness of hypoglycemia, citing a 72% reduction in hypoglycemic events in type 1 MDIs in the HypoDE study.
- (v) CGM + Pump to Hybrid closed loop. Commenting that during his fellowship, where he worked on the so-called “Blue Brick” auto-syringe, he never thought he’d see automated insulin delivery in his career, Dr. Bergenstal discussed the advantages of the MiniMed 670G system. He discussed pivotal trial data in adults and pediatrics (7-13 years), the overnight benefit (“There is no such thing as the right basal rate overnight”), and some practical tips, including: run manual mode ≥6-7 days prior to initiating auto mode; stay in auto mode early and often (>80%); strengthen insulin:carb ratio 10%-20%; if high/low 2-3 hours after a correction bolus, shorten/lengthen active insulin time; <15 grams of carbs are often needed to treat mild lows because basal rate has been diminished/shut off for some time already; and start temp target 1-2 hours before exercise, end temp target 1-2 hours after exercise. He emphasized that 670G is a lot of work – as evidenced by his tips for success – but videos of 670G pivotal study patients showed they felt it was worth the effort.
- The slide below depicts the remarkable results from one of Dr. Bergenstal’s long-time patients – a DCCT participant – on 670G. There is unfortunately no “before” period for comparison, but it doesn’t get much better than 94% in-range with 2% time <70 mg/dl and 4% >180 mg/dl. Prior to starting on 670G, the woman was skeptical that the system could do any better than one of the world’s top endos had done for the past 35 years.
- (vi) You can do it to You and your NEIGHBORHOOD can do it. It’s not just about the patient, but also about the support they get from HCPs, caregivers, and technology. Setting the tone for this last major transformation, Dr. Bergenstal casually strolled to the podium (against his wife’s warning that it was silly) to begin his talk with his blazer draped over his shoulder to Mr. Rogers’ “It’s a Beautiful Day in the Neighborhood.” The words “Dr. Bergenstal’s Neighborhood” were displayed on a slide, and the audience rewarded the show with laughter and applause.
In a rousing morning address, Sansum’s Dr. David Kerr delved into the role of digital health in modernizing diabetes (a.k.a., addressing the myriad determinants of health), detailing some fascinating efforts from his center in food as medicine, environmental determinants, and leveraging consumer apps for clinical research. Prior to jumping into the specifics of each project, Dr. Kerr took the liberty of summarizing the whole talk, borrowing a quote from Sir William Osler: “It is much more important to know what sort of patient has a disease than what sort of disease a patient has.” He also pointed out that this is a “pharmaceutical-free zone,” that modernizing diabetes care doesn’t need to rely on medications (beyond insulin); of course, as he pointed out, his perspective for the purpose of this talk was “biased,” and there is a role for pharmaceuticals in tandem with systems-level and digital interventions in addressing diabetes. Read about three of Dr. Kerr’s ambitious projects to deliver parity in diabetes outcomes in Santa Barbara immediately below.
- Dr. Kerr was particularly excited to share encouraging pilot data from the Farming for Life program, a “food as medicine” initiative for some Latino Santa Barbara natives with or at risk of type 2 diabetes. He and his team surveyed members of this community to find out the top five vegetables they would eat, then went to local farmers and food distributors to strike a deal. Then, a doctor “with a white coat and a stethoscope” saw the person and wrote a medical prescription for vegetables. Without even giving the participants instructions on what to do with their vegetables, a small pilot (n=25) yielded slimmer waistlines and a ~12 mmHg drop in systolic blood pressure. For this simple concept, Sansum has received a federal grant to scale to 400 Latino individuals!
- For the Mil Familias project, Sansum has recruited (with the help of Latino community health workers dubbed “Especialistas”) 1,000 Latino families to partake in studies interrogating why it is that Mexican Latinos have a type 2 diabetes prevalence of 18%. These families have agreed to allow Sansum to study aspects of their determinants of health – genetics, biology, behavior, psychology, and environment – and even partake in interventional studies (no specifics here). In this very early stage, Sansum has already discovered that although these people are at risk of diabetes (or have type 2 diabetes), they are fairly active, taking a median of >7,000 steps per weekday, likely to keep up with the demands of multiple jobs. A side nugget from this data is that Fitbit consistently over-reports step count by a magnitude of ~2,000 steps per day vs. the gold standard “ActiGraph,” a very wise move on Fitbit’s part, if intentional, to make users feel good about themselves and therefore the product stickier. The discrepancy could also be attributed to specific activities commonly seen in this population that fool the Fitbit but not ActiGraph into registering steps. Dr. Kerr et. al have big plans for Mil Familias, including a deep dive on the participants’ environments – particulate matter, temperature, noise, light, barometric pressure – complete with sensing devices that can go “in neighborhoods, on homes, in homes, on clothes…we’re even having a conversation with the vaping industry to see if we can change vapes into sensors that can measure microparticles in the breath.” The idea is to zero in on factors that could put certain individuals or communities at higher risk of developing diabetes.
- Noting that most Latino families don’t have a car, Sansum is looking into contracting with rideshare apps (Uber, Lyft) to bring participants to and from clinic visits. Dr. Kerr offered the example of a person who has a two-bus, 45-minute commute to and from the center. Since the cost of a participant not showing up to an appointment or cancelling at the last hour is nontrivial, it makes perfect sense for Sansum to invest in rides. We’ve heard of rideshare contracts for making medical appointments before, but never in the context of clinical research!
- “Clinicians are obsessed with glucose. But the people I see in clinic are obsessed with time the time they have to spend on diabetes self-management. We’ve forgotten in this community to think of time as a currency. If you want to modernize diabetes care, improve glucose and blood pressure, but also a key metric of success is reducing the time someone with the diabetes spends working on the disease.” We love this sentiment! Of course, the business case seems harder to make, though a patient spending less time on diabetes is likely doing so because he is spending more time-in-range or is now more willing to focus on trouble spots in his diabetes management because other portions of his time have been liberated. Time is a wholly undervalued resource in diabetes!
- Work from Evidation Health and Sansum (figure below) demonstrates the power of big data: Sleep quality begins to deteriorate and activity begins to decrease ~five days before the onset of the flu. There’s no diabetes-specific learning in this data, but it does show how tracking various phenotypic parameters can shed light on early indicators and lead to early intervention (flu treatments are far more efficacious when commenced before flu onset). This is similar to the guiding principle behind One Drop’s and Medtronic’s efforts predict blood glucose up to 12 and four hours in advance, respectively.
- Dr. Kerr explained that apps, connected pens, AI/big data, and blockchain will transform diabetes:
- Citing a May 2018 review, he conceded that of the 318,000 diabetes apps out there in 2017, only 11 had been evaluated in outcomes studies. Five showed a clinically significant improvement in A1c, but none improved quality of life, blood pressure, or weight, and due to “methodological issues,” none were considered to be of high quality. A recent meta-analysis, however, show that apps involving HCP feedback confer significantly greater A1c reductions than those that don’t (-1.12% with high HCP feedback vs. -0.24% with no HCP feedback), and we’d note there are countless observational studies (some prospective) showing a benefit to diabetes app use.
- Dr. Kerr asserted that smart pens – specifically referencing Novo Nordisk’s recent smart pen announcement – are going to change the working lives of clinicians looking after people with diabetes. He referenced an interesting JDST article he co-authored with Dr. David Klonoff, which includes a run-down of clinical settings in which connected dose capture device will likely prove beneficial.
- Dr. Kerr warned attendees to be wary of AI, but also touted its ability to predict outcomes, make decisions for people, and convey meaning and emotion (“What a great week, all of your blood sugars have been in range, fantastic!”). Despite the great potential for AI to impact diabetes outcomes positively, he preached “humility not hubris” as the field approaches it, conveying his message with a cautionary tale from the Vietnam War. Former Secretary of Defense Robert McNamara fell victim to a “quantitative fallacy” in the war, believing that he could use data and math to plan and win a war. The fallacy is this belief, that if you can measure something and put it in an algorithm, you can achieve a desired outcome, but that’s not how the Vietnam War turned out in the end. A quantitative fallacy is committed when the actor: (i) measures whatever can be easily measured; (ii) disregards that which can’t be measured or gives it an arbitrary quantitative value; (iii) presumes that what cannot be measured easily is not important; and (iv) believes that what can’t be easily measured really doesn’t exist. When AI software makes predictions or therapeutic decisions for people with diabetes, then it is clearly paramount that all relevant inputs are identified and somehow incorporated into the model.
- We were highly intrigued by Dr. Kerr’s ambitious vision of a world where Blockchain enables patients to get paid for their health data (despite his caveat that “it might fall over”). In the seemingly well-thought-out scenario, patients would provide data, which would then be held in a distributed fashion by multiple stakeholder (via Blockchain). Only the patient would hold the key to the data, such that every time someone wants to access or use the data, the person with diabetes gets paid or otherwise rewarded. For example, if they provide blood pressure data and a research organization uses it, the patient will receive “10 credits,” which they can then use to buy vegetables or contribute to their cell phone bill, etc. (ideally, the credits would be tagged for health-promoting behaviors). Further, if the patient not only provides blood pressure data but demonstrates a reduction, then they could receive “100 credits." The data would go nowhere near the EHR unless the patient wants it to, and Dr. Kerr’s radical vision is that one day EHR vendors will actually be paying patients for their data!
Joslin Diabetes Center’s Dr. Medha Munshi discussed the unique challenges of and management strategies for managing type 1 diabetes in the elderly. With an aging population generally and people with type 1 diabetes specifically living longer, this topic doesn’t receive nearly enough attention – Dr. Munshi opened the talk by commenting that she was glad to see people in the room, because she feels “lonely doing this in the US.” Reviewing data from ADA 2018 that compared the disease profiles of older individuals with type 1 diabetes, older individuals with type 2 diabetes, and younger individuals with type 1 diabetes, Dr. Munshi concluded that aged type 1s are at a particularly high risk of complications. In the presented cohort (n=54), older type 1s were on an average of 10 total daily medications, had five comorbidities, and 17% had a recent fall (these numbers in older type 2s were slightly higher, and they wre significantly lower in younger type 1s). Similarly, older type 1s had a high prevalence of cognitive dysfunction (57%), vision impairment (19%), and hearing impairment (33%) – all of which can impact ability to care for oneself. Older type 1s had higher hypoglycemia risk than older type 2s (similar to that of type 1s), and easily had the highest hypoglycemia fear scores (mean 40) and highest reported impaired awareness of hypoglycemia (44%). Simply put: These individuals are at high risk of complications and may not be the easiest to treat.
Accordingly, Dr. Munshi strongly recommended making therapeutic changes in small steps and simplifying care regimens to adapt to self-care barriers, caregiver ability/stress, and patient preference. She gave three examples of this simplification strategy: (i) She put an 82-year-old woman on basal-bolus insulin with three different scales and highly variable blood glucoses on a simplified regimen of basal in the morning and fixed dose mealtime insulin (4-3-3, with adjustments for abnormally large/small meals). “No problem solving, no complex scale, and no need for fingersticks before each meal.” (ii) A 90-year-old woman who lives with a husband with dementia and is starting to be forgetful, is on a basal-bolus regimen (suspected missed doses), and whose daughter is around in the morning/at night but works during the day. Dr. Munshi put her on basal and mixed-dose insulin in the morning – administered by the daughter – to provide all-day coverage, with the option for insulin before meals and for corrections. (iii) Dr. Munshi helped an elderly man figure out his 2-5 most common meals and identified insulin doses to match each one – no calculations necessary! The underlying message of simplification, which she tells patients and their caregivers, is that it is definitely not the best way to treat diabetes by far, but it is the best way to treat you. We love this philosophy; an overly-complex but optimized regimen is useless if it’s not followed. One thing Dr. Munshi didn’t touch on is the role of technology (closed loop, decision support) in this population, which we’re particularly curious – which technologies does she think would be particularly useful now, and what coming down the pike (photo meal recognition and bolus calculation?) would she find particularly applicable for treating diabetes in the elderly type 1? She stressed that she doesn’t endorse liberalizing goals, letting the A1c run high or de-intensifying therapy, but a “very deliberate way of changing the strategy of giving insulin based on where you see barriers.” Nice! She also emphasized very intensive diabetes management when patients are younger and in middle age so that they are in better shape when they approach old age. She concluded by reciting the Serenity Prayer (of Geriatrics): “Grant me the serenity to accept the things I cannot change, courage to change the things I can, and wisdom to know the difference.” In case there was any doubt as to what she meant, the next slide had a picture of an older woman lighting a cigarette from the “0” candle on her 100th birthday cake.
- The diabetes management strategy in a given elderly type 1 must be heavily dependent on the living situation. Dr. Munshi explained that a person living independently at home may have a hard time with complex medication regimen, particularly if an acute illness causes a decrease in cognitive or physical status, and would likely need frequent education and re-education. For someone living in an assisted care facility, the provider may have no control over meal content, the patient will likely have assistance with medications but not blood glucose monitoring/insulin administration, and there’s a high risk of failure after an acute illness. Finally, for a person living in a nursing home, the HCP has no control over time/content of diet and there’s a higher risk of side effects from oral medications and higher risk of acute illness, anorexia, dementia, and delirium, but on the plus side, self-care activities are performed by the nursing home staff (it’s possible to get away with a more complex care regimen).
- Though some elderly patients appear frail and others seem more able and capable of handling more complex/intense regimens, Dr. Munshi pointed out that the age number still makes a difference. This is based on the principle of homeostenosis, which is the progressive decline in humans’ ability to maintain homeostasis in the presence of environmental, physical, or emotional stress. No matter someone’s degree of frailty, their ability to bounce back from a stressor decreases over time, and this can lead to poorer outcomes if confronted with a stressful/over-intense management strategy at too advanced an age.
LMC Diabetes and Endocrinology CMO Dr. Ronnie Aronson delivered what Ascensia coined “not another boring BGM talk,” discussing the comparative accuracy of CGM and BGM, as well as CGM calibration. Following the talk, we heard more than one attendee exclaim, “Ohh I finally get MARD” – Dr. Aronson is a master of simplifying complex concepts.
- Dr. Aronson shared a helpful rule of thumb for cross-comparing CGM and BGM accuracy: If a BGM meets ISO 2013 guidelines of 95% values within ±15% of reference, MARD will typically be ~6%. Generally, divide the “width” of the bounds (±20%, ±15%, etc.) by 2.5 to arrive at the MARD. Citing a seminal 2010 UVA study, Dr. Aronson pointed out that there is now actually a push toward ISO standards of ±10%, since in silico modeling shows this would limit missed hypoglycemic events to just 1 in 100 (vs. 4 in 100 missed at 15% BGM error and 10 in 100 missed at 20% BGM error). The Ascensia Contour Next was the one BGM of 18 tested in the DTS BGM Surveillance Program to have 95%+ of trials within ±10% of reference. Meanwhile, Dr. Aronson displayed data from an LMC study showing that Senseonics’ Eversense XL 180-day implantable CGM only meets the 95% criteria when bounds are expanded to ±30% – at ±20% of reference, only 92% of values fall in range. Concluded Dr. Aronson, “When you’re hearing about the latest and greatest CGM, they’re trying to get below 10% MARD, they’re still nowhere near what BGM can do in terms of accuracy. If we are to meet ISO requirements for BGM, that is the kind of standard that we might want to apply to CGM. A goal for the future.” At the same time, he preceded the discussion on BGM accuracy by sharing that many of his patients dose for frequent corrections based on directional CGM arrows, rather than absolute values. Indeed, BGM may have greater point accuracy and no lag time, but the value of real-time, non-invasively retrieved, trending glucose data provided by CGM can’t be overlooked.
- Dr. Aronson mentioned during Q&A that many of his patients are actually performing more frequent SMBGs when they are using FreeStyle Libre vs. when they are on SMBG alone. “I’m finding that my patients need to use SMBG at the lower end of the glucose range, where they’re getting either incorrect low readings, or feeling low and the sensor is telling them they’re at 4.5 mmol/l (~81 mg/dl).” He later clarified that those patients of his who try Libre are not quite representative, as they are likely to try to run in the low end of the target range, and they’ve been getting both false lows and false negatives. “The result is that when people are anxious, they test more.” The 14-day, real-time, FreeStyle Libre’s labeled accuracy in hypoglycemia has lagged behind other systems, particularly in that it has concerningly high false hypoglycemia notifications (between 52%-72%, depending on the method). We frequently hear comments that Libre over-reports hypoglycemia, though Dr. Aronson’s account of under-reporting is rather new. Still, SMBG frequency went way down when patients used FreeStyle Libre in IMPACT, REPLACE, and the real-world, so we assume the higher observed frequency in Dr. Aronson’s patients is related to the specific, uber-user population.
- Ascensia’s Dr. Andreas Stuhr likened CGM calibration to wearing a seatbelt: “Why wouldn’t you use the seatbelt in the car to prevent an accident? It doesn’t mean you will have an accident when you put the seatbelt on, but there are situations where CGM is just not perfect yet, just like BGM wasn’t perfect at the beginning. So why not use a seatbelt twice a day?” This is a very interesting analogy, but we’d caution that calibrations can sometimes introduce error into CGM measurement when technique is lacking; so, to adjust the seatbelt metaphor accordingly, imagine a scenario where it is possible to fasten a seatbelt so tightly that it impairs the ability to safely drive. With the G6, we’ve heard anecdotally from many people that they never calibrate, while others still insist on doing so once or twice a day – we’d be curious to see data from Dexcom outlining users’ calibration behavior now that it is optional.
Just two days after the MiniMed 670G was licensed by Health Canada, Yale’s Dr. Jen Sherr and Northwestern’s Dr. Grazia Aleppo presided over an upbeat Medtronic dinner symposium. The first units are expected to ship in Canada in December, so attendees were eager to soak up as much knowledge on best practices from the two experienced speakers as possible. The energy in the room evoked a strong sense of déjà vu – hearkening back to the days in proximity to the US – as did a number of the comments we heard from nurses and educators in the room: “I have a friend in Indianapolis who loves her pump, I just can’t wait to get my hands on one.” “What am I going to do with all of my free time when I start wearing the 670G?” “How does it change insulin rates? I’m a little scared to just put it on and let it go.” A Medtronic rep introducing the session addressed this last question, drawing a parallel to self-driving cars. The first time he saw automation in a car, he said, was when his friend’s Tesla parked itself. “My eyes were wide, I was anxious, just praying we wouldn’t hit the Mercedes behind us. But it was fine.” Drs. Sherr and Aleppo clearly explained best practices we’ve heard so much about in the past year-and-a-half; the rest of the world would’ve liked to have 670G sooner, but they will certainly benefit from the stateside learnings. Below, we’ve compiled some of our favorite quotable quotes from the talks.
- (On the case of an 8-year-old type 1 with frequent and severe hypoglycemia who started on 670G) “This mother woke up every two to three hours every night, and wouldn’t give the full bolus because she was scared of lows… We transitioned her to auto mode, and we didn’t make huge changes to settings, we just finally let the system work. Overnight, this girl is now in the target range. Mom looked like a new woman, because she had slept. There’s a reason infancy is a year because we can’t handle a lifetime of that. This woman did it for four years. Now she goes to sleep and knows her child is safe.” – Dr. Sherr
- “This technology has actually shown my patients that they have to bolus before meals. I’ve said it for years, this technology taught them in a week.” – Dr. Sherr
- “Just remember the numbers 90, 80, and 70. At least 90% sensor wear, 80% time in auto mode, and 70% time-in-range. That’s the basic thing to remember.” – Dr. Aleppo
- “By having to do less of the work, I have better conversations, which is why I wanted to work with patients in the first place.” – Dr. Sherr
- “We have over-basalized our patients for ages, trying to overcome inadequacies of bolus insulin. 50:50 bolus:basal ratio is no longer correct. It never was, but this system will show us that’s absolutely the case. 40%-45% basal is what you want.” – Dr. Aleppo
- “In the pivotal study, the oval tape (for the Guardian Sensor 3) was like gold. If you have patients who are wearing sensors, they want the full duration. The oval tape has really overcome that for many patients, so that they’re now confident they’ll get the full seven-day wear.” – Dr. Sherr
- “Teach your patient not to fight the algorithm. If you fight you lose. If you work with the system, and bolus appropriately, it will help you tremendously and keep you in Auto Mode” – Dr. Aleppo
- “Some patients who went on the 670G were using other sensors at first. They needed to know they could trust this new system. One of the patients kept wearing her other CGM. But three days into wearing the system, her mom told me she had taken it off and threw it out.” – Dr. Sherr
- “It’s important that when the JDRF CGM trial was done years ago, by the end of the study, only 30% of the participants were wearing the device six days a week. In the 670G pivotal, auto mode utilization was 88% in adults, 76% in adolescents, and 75% in children. I have teens, believe me, parents aren’t going to be able to make their kids wear it, they’re going to have to want to. So I found that very impressive.” – Dr. Sherr
- “The CareLink reports will tell you exactly what the patient is doing and how to improve time in range.” – Dr. Aleppo
Obesity, Psychosocial, and Big Picture Highlights
In a truly engaging keynote, Dr. Sean Wharton (Wharton Medical Clinic, Hamilton, Ontario) dialed up several calls to action as the world faces an incoming “obesity tsunami.” He showed off some serious moves by dancing on stage to “Purple Haze,” and didn’t slow down for the rest of the talk, providing an extensive overview of preconceptions regarding the pathophysiology of weight change, current treatment options, and emerging therapies in obesity medicine. Dr. Wharton’s talk was also marked by several calls to action in order to effectively address obesity. Most notably, he urged the audience to reevaluate their biases against medication in obesity therapy. According to Dr. Wharton, pharmacotherapy is the only effective and scalable current option for obesity management. Surgery has proven effective in some cases, but is clearly not scalable at a population level. Similarly, lifestyle interventions, while assuredly essential as a first step, show limits to their efficacy, and are not entirely scalable – Dr. Wharton commented that most of the benefits seen with these modifications are in very rigid and resource-rich environments. Moreover, lifestyle interventions are often predicated on willpower from the patient, and Dr. Wharton contends that this is an exhaustible resource that is not sustainable and does not last past the timescale of months. Pharmacotherapy stands apart in balancing efficacy and scalability. However, medications in this arena remain woefully underused: Dr. Wharton shared the shocking statistic that only two percent of patients with obesity are treated with an anti-obesity medication. This number is incredibly low, and Dr. Wharton attributed it largely to bias among all stakeholders involved against the idea of using medications to treat obesity. He characterized obesity as a “chronic, progressive, relapsing medical condition,” and just like any other condition of this nature, it must be treated consistently with pharmacological agents in order to improve. Taking a step back, we do wonder how much of this underutilization is due to bias, and how much is due to other factors such as a lack of truly bona fide and efficacious obesity drugs and high near-term costs associated with existing therapies (although it is certainly possible that bias mediates both of these variables as well!). For more on the status of obesity pharmacotherapy, see our simultaneous coverage of the Cleveland Clinic Obesity Summit, where we heard a comprehensive overview of all FDA-approved obesity therapies.
“We need to move past diet wars.” On the subject of the seemingly endless iterations of diets recommended for obesity treatment – including keto, paleo, vegetarian, vegan, Mediterranean, raw, law carb, and no sugar, to name a few –Dr. Wharton cheekily proclaimed, “I believe in all of these diets.” He contended that he’s happy with his patients going on any reasonable diet that they will actually stick to, and that a better diet doesn’t count for anything unless the patient actually follows it. On this point, he emphasized again that willpower is an exhaustible resource that does not last; therefore, patients need to like the diets that they are on for them to actually stay on them for a longer period of time and show some level of success.
Dr. Wharton emphasized that weight loss is an inherently heterogeneous process, so averages don’t tell the whole story! It’s a well-established fact at this point that long-term maintenance of weight reduction is extremely difficult. With lifestyle interventions, meta-analyses have shown that significant weight loss is possible in the first six months of treatment; however, over the subsequent years almost all of this lost weight is re-gained. Dr. Wharton’s clinic sought to see if variability existed in this data from patient to patient. In other words, were some patients able to keep off the weight loss entirely over time? The clinic collected data on 29,052 patients, and similarly saw that on average, patients using lifestyle modifications achieved initial body weight losses of 2.9%, but quickly started to regain significant portions of this amount within two years. Tracking the individual trajectories of patients’ body weight loss percentages over time yielded a different story (in another group of ~7,000 patients over 7 years who were using lifestyle modifications), however (graph below). Different patterns for weight loss over time emerged. Nearly two-thirds of patients were characterized as “weight stable” (purple line) or “minimal weight loss” (yellow line). However, Dr. Wharton highlighted that some patients did show large weight loss effects sustained over time (2.3% of patients that exhibited on average a 21% weight loss). Dr. Wharton conceded that he had no idea why some patients were more successful with fending off weight regain, but the results establish that weight loss maintenance over time is possible. The logical next step remains determining what’s different in these patients – be it behavior, medication, age, or something else – hoping that the difference is modifiable, and applying these learnings to a broader population.
Drs. Noah Ivers (University of Toronto), Robyn Houlden (Queen’s University), and Ms. Susie Jin (independent pharmacist) overviewed the 2018 Diabetes Canada guidelines, placing a special emphasis on the intersection of mental health effects in diabetes management. Three different “challenging” clinical situations were overviewed as a window into how to effectively use the guidelines and the many resources that come along with their development. Notably, the Diabetes Canada guidelines are in strong agreement with those from ADA/EASD. In the session, we greatly appreciated the discussion on how to effectively consider mental health aspects of diabetes care, an oft-overlooked comorbidity.
- A brief comment during Q&A contextualized these broader guidelines, as Dr. Robyn Houlden joked that “we have a strong feeling that they [ADA/EASD consensus guidelines] have copied our guidelines.” The ADA/EASD consensus guidelines were just recently released during EASD 2018; the Diabetes Canada 2018 guidelines were released earlier this year in April. Dr. Houlden noted that the pharmacological algorithms used in both guidelines are extremely similar – we’re glad to see high consistency across these guidelines. Below, we’ve shown the treatment algorithm from the 2018 Diabetes Canada guidelines, which are quite similar to the ADA/EASD consensus guidelines. However, we do note that the ADA/EASD guidelines do provide more granularity when considering the presence of clinical CVD, as more direction is provided in looking at the contributions of ASCVD, heart failure, and CKD (see our full coverage of these guidelines here). For example, the ADA/EASD algorithm specifically recommends either an SGLT-2 or GLP-1 with proven CV benefit for those with predominant ASCVD, but prefers an SGLT-2 over a GLP-1 if heart failure or CKD predominates (SGLT-2s should only be used if eGFR is adequate).
- An individual patient case was presented to illustrate the importance of addressing barriers to treatment, especially those involving comorbid mental illness. The patient, Keisha, is a 55-year-old woman with a history of major depressive disorder and an A1c of 7.7% at initial screening. She receives the whole armory of standard care off the bat: she’s started on metformin, receives individualized nutritional counseling with a registered dietician, and is seen by a diabetes educator who reviews healthy behavior interventions (including physical activity recommendations), antihyperglycemic medication adherence, and the self-monitoring of blood glucose. Keisha is even given a “smarter step count prescription” to help her increase her baseline steps by 3000 per day within a year. By all means, Keisha has had tremendous resources in terms of care and attention directed toward her. However, after eight weeks, her A1c has not improved, and Keisha returns to her follow-up meeting in tears. Even with these proposed interventions and steps of treatment, she says that she’s “feeling worse than ever” and that she “feels overwhelmed at home and work.” At this point, the speakers recommended that Keisha take the diabetes distress screening questionnaire. They emphasized the very strong link between the diabetes and mental illness: Comorbid mental illness is known to lead to non-adherence with medication and self-care, along with increasing the risk of complications; on the other end, major depressive disorder increases the risk of type 2 diabetes by 60%. We do wonder if it’s even worth Keisha’s time to take the questionnaire – there’s clearly an issue, regardless of whether it is related to therapy or mental health and coping with diabetes and its management.
- One tool emphasized in the 2018 Diabetes Canada guidelines is the well-known “diabetes distress scale.” HCPs can access the resource in the “Health-Care Provider Tools” section of the Diabetes Canada guidelines. This scale is meant to screen patients for high levels of distress, and contains a series of questions for the patient to answer on potential sources of distress: patient and provider, the emotional burden of living with diabetes, the burden of self-management, and burden related to social aspects of living with diabetes. Patients with a score of three or higher on the scale should be encouraged to complete the full diabetes distress screening questionnaire. HCPs should attempt to identify whether a patient has diabetes specific distress (upon which diabetes care and support management should be directed to the patient), a mental health disorder (upon which mental health professionals may need to be involved), or living distress (upon which patients may need to be helped with stress management).
University of Alberta’s Dr. Jane Yardley gave a primer on exercising with type 1 diabetes (Insulin, Intensity, and Insulinization). In the course of her talk, she raised some unanswered questions and under-appreciated exercise quirks (impact of time of day!), speculated as to why studies of aerobic exercises often don’t show A1c benefit, and shared new data highlighting the importance of early insulin adjustment before aerobic exercise (for pumpers).
- According to Dr. Yardley, the vast majority of studies to date have been in young fit males – how does the blood sugar of female type 1s respond to different forms of exercise? Based on the seed observation that women report more hyperglycemia during the luteal phase of menstruation, Dr. Yardley is investigating the impact on exercise (does the patient require more insulin?). In addition, estrogen alters the body’s fuel preferences, perhaps such more lipids are used up while carbohydrates are spared, implying that the type 1 woman may be at lower risk of hypoglycemia during exercise than the type 1 male. “We have no idea,” she said, “I’m working on that one.”
- Dr. Yardley hypothesizes that overcompensation – heading into exercise above target to prevent hyperglycemia – is one of the reasons studies of aerobic exercise often do not find improvements in A1c. She showed telling figures from her 2013 study where type 1s were instructed to come in to clinic and either sit idly or to perform an exercise – while all individuals performed each task, they didn't know whether they would be exercising or lounging on a given day. Despite being asked to repeat food intake and insulin adjustments exactly between sessions, CGM traces from two of the individuals clearly indicate that last-minute adjustments were made when they found out they would be performing aerobic exercise (elevating sensor glucose level into the 220-290 mg/dl range). Dr. Yardley noted that the benefits of exercise are “probably more beneficial than observed” (as dictated by A1c), but the striking data show that there is plenty of room for therapeutic improvement (or at least there was as of 2013). There is a clear role for hybrid closed loop and PLGS to play in providing patients a “safety net” as they approach the daunting task of managing diabetes while exercising.
- Why important to reduce insulin in advance. (2018 data). New data Roy-Fleming 2018 in press. Why it’s important to reduce insulin well in advance. Early adjustment really important for people on pumps. Survival curve – can go longer before supplementing with carbs.
- Surfacing an oldie but goodie, Dr. Yardley reminded attendees that the time of day drastically changes the body’s response to diabetes. Exercising first thing in the morning before breakfast typically causes blood glucose to surge. Researchers aren’t entirely sure why this happens, but it could involve lower circulating insulin and higher levels of growth hormone and cortisol. Physiological explanation aside, this finding means that basal adjustments should be “thrown out the window” when exercising first thing in the morning before eating, in many people.
University of Saskatchewan’s Dr. Terra Arnason and University of Calgary’s Dr. David Lau faced off in a debate over the addition of intermittent fasting to a type 2 diabetes care regimen. Moderator Dr. Alice Cheng recognized Dr. Arnason, arguing the “pro” side, as the winner at the end, but in reality, picking a champion was complex. In a show of hands, we estimate 75% of the audience voted “pro” prior to the debate, but voter turnout was abysmally low. In the post-vote, the split appeared to be 50:50, but there was a much bigger voter turnout, and the vast majority of people who switched sides did so in favor of “pro.” The message behind the muddle is that the jury is still somewhat out, and fittingly, both debaters agreed that long-term studies are lacking and needed. Dr. Arnason methodically checked off three boxes – simple, healthy, and effective – in her opening statement, leaning mostly on small, short studies, many of which were not in people with type 2 diabetes. Dr. Lau addressed the same points, often drawing different conclusions from the same studies, before ultimately claiming that intermittent fasting is “plain hype and not yet ready for prime time.”
- Dr. Arnason on simplicity: Dr. Arnason posited that all varieties of intermittent fasting are very simple, drawing clear boundaries and allowing the individual to never miss an evening meal with family. We found her argument that the 5:2 variety of intermittent fasting (two days per week fasting interspersed throughout each week) entails 104 days of dieting and 200+ days of ad lib feeding to be particular compelling; traditional caloric restriction entails 365 days of willpower. In her analysis, multiple clinical trials reported >75%-85% compliance with intermittent fasting throughout study periods of three months to a year, and patients reported that they would continue after.
- Dr. Lau: Pointing to Dr. Jason Fung’s guide book on intermittent fasting, Dr. Lau asked, “is it really simple? If it’s simple, you don’t need a guide.” To his point, dropout was fairly high in most of the studies presented, though rates were similar to those seen in caloric restriction comparator arms.
- Dr. Arnason on healthy (+safety): Dr. Arnason challenged the audience to consider why eating three meals based on time of day rather than in response to hunger is healthy. Going back in time to our hunter/gatherer roots, it was the norm to eat several days consecutively after a successful hunt, but otherwise it was a near-complete feast. Today, food is constantly available, and we rarely stop eating long enough to truly feel hungry or move into a net-negative energy balance; three spaced meals plus snacks maintains an anabolic state, which Dr. Arnason hypothesized may contribute to the “weight loss plateau” or unexpectedly low impact of exercise on weight loss. She added that short-term fasting stimulates fasting (+5% increase in resting energy expenditure vs. two weeks of continuous caloric restriction reducing resting energy expenditure by 10%) and has benefits beyond weight loss (consumes fat stores, decreases insulin resistance, etc.). Moving to safety, Dr. Arnason first noted that millions of people fast occasionally for religious purposes, many of whom statistically have diabetes. One big question mark concerning intermittent fasting in type 2 diabetes is around hypoglycemia (insulin/SUs + no food could equal bad news). Dr. Arnason made the case that the risk is not large: In Carter et al., there was no difference in hypoglycemia levels after one year between the intermittent fasting and caloric restriction groups, though 46% of participants overall used SU or insulin, and 20% in both arms used insulin. There was no severe hypoglycemia reported, and the study protocol called for insulin/SUs to be pre-emptively decreased or stopped based on A1c at entry and depending if it was a fasting day.
- Dr. Lau: Dr. Lau was generally skeptical about all of the promised benefits of intermittent fasting, but his strongest argument came from appropriating a paper Dr. Armason earlier cited as evidence of safety (Corley et al.). While the study, whih was rich with insulin users, showed lower overall hypoglycemia in intermittent fasting, Dr. Lau pointed out that risk of hypoglycemia was two-fold greater on fasting days (RR 2.05). This heightened risk came, he said, despite weekly supervision, hypoglycemia education, and structured medication reduction – clearly not a non-labor-intensive proposition. Dr. Lau came back to this point many times throughout his spots: intermittent fasting may be safe and beneficial for some people, but it seems to require a lot of oversight. He also cautioned the risk of ketosis and even ketoacidosis, especially in people who are on SGLT-2 inhibitors. Lastly, he suggested that fasting can lead to adverse effects such as headaches, decreased energy level, constipation, diarrhea, feeling cold, lack of concentration, mood swings, and preoccupation with food (Dr. Armason suggested these could happen in any dieting plan).
- Dr. Arnason on effectiveness: A three-month trial comparing intermittent fasting with 25% daily calorie reduction (Carter et al., 2016) found equal weight loss (-5.9%), A1c drop (-0.7%), appetite score changes (increased feeling of fullness and satisfaction), and compliance in both arms – Dr. Arnason reminded that she was tasked with proving effectiveness, not superiority. The same study design run out to a full year (Carter et al., 2018) yielded more of the same, with non-significant differences in A1c drop or weight loss. Recent studies with alternate day fasting in non-type 2 people with obesity showed decent weight loss and excellent adherence at 11 weeks (Varady et al.) and equivalent weight loss and maintenance out to one year when compared to caloric restriction (Trepanowski et al.) Finally, she pointed to Furmli et al., a three-person case study published last week in which three people on 70 units of insulin daily managed to cease insulin therapy, maintain blood glucose, lose weight, and reduce waist circumference through alternate day fasting.
- Dr. Lau: Dr. Lau made the case for a lack of superiority, but did not address non-inferiority to caloric restriction. Bowen et al. showed no significant difference in weight loss between 16 weeks of alternate day fasting and daily energy restriction (as a side note, quality of life measures were lower in intermittent fasting arm); same story with Trepanowski et al. and Carter et al. (2018), though he added that the latter missed its primary endpoint of A1c reduction, and A1c even trended higher in the alternate day fasting group.
After serenading attendees with a rendition of Johnny Cash’s “Ring of Fire,” famed country singer Mr. George Canyon suggested adopting a “corporation model” in adolescent diabetes care, where the teen is the CEO. In this paradigm, the corporation (caregivers and healthcare providers) have to explain to the teen what a CEO’s responsibilities are, but they must keep in mind that they work for her at the end of the day. If the parents are too aggressive, hounding her for not bolusing before a meal, she can “fire” them. Fire, in this case, means to shut out: “They can start to lie through their teeth about their diabetes and what they’re doing about it and there’s nothing their parents can do.” But at the end of the day, the CEO also has to respect what her employees are saying, because otherwise there’s no company and she’ll end up the “CEO of nothing.” The analogy is another way of saying “patient-centered care,” but adds a tangible and self-important aspect that all parties can glom onto. Mr. Canyon says he’s seen this model work in the past, opening up lines of (honest) communication between adolescents and their parents.
- Should we change the name of type 1 or type 2 diabetes to something else so that they are no longer confused by the general public? Admitting that he is strictly coming from the patient perspective, Mr. Canyon was adamant that such a change is needed, offering to lend his “big mouth” to any efforts already underway to do so. In his countless hours talking with children with type 1 diabetes, he has repeatedly heard stories of misunderstanding leading to stigmatization. For example, a little boy’s schoolmates told him the only reason he has type 1 diabetes is because he wants attention – all he has to do is take is medication, exercise, and eat right and he won’t have diabetes anymore. “Kids plead with me all the time. They know this is wrong. We need to change the name of the disease.” Mr. Canyon’s point is somewhat interesting although it also completely over-simplifies type 2 diabetes in our view. On the flip side, one could argue that the underlying pathology of type 1 and type 2 diabetes is related enough to warrant similar names (though one could also take the other side of this argument). One of the best arguments for keeping the current nomenclature is based in pure pragmatism: Diabetes research and intervention receive very little funding relative to other disease areas when adjusted for scale of impact. While the type 1 community is more vociferous today than the type 2 community, it has 20-times fewer members than does type 2. Combining the type 1 voice (while de-stigmatizing and encouraging the type 2 voice) and the scale of type 2 diabetes may be a potent stimulator of investment in both types of diabetes. Banding together in advocacy efforts may be the most practical step for drawing investment, and it will also raise awareness of both diseases, which could prevent additional type 1s from being teased for “not exercising enough.”
McGill’s Dr. Kaberi Dasgupta shared some disheartening data about the link between stigma and glycemic outcomes. Of 178 teens and 202 young adults with type 1 diabetes reached through and online survey (380 people total), 66% reported experiencing some degree of diabetes stigma. The proportion who said “yes” was slightly higher in young adults (69%) than in teens (61%). Those who reported experiencing stigma were twice as likely have a high A1c (>9%) or a severe hypoglycemia in the past year; drilling down into specifics, they were three-times as likely have a high A1c and twice as likely to have a severe hypoglycemia event in the past year. It’s not difficult to imagine how feelings of shame and judgement could lead to delayed treatment, missed boluses, and a general desire to avoid engaging with diabetes in public. It’s also possible that worse diabetes outcomes beget perceptions of stigma, but there is clearly a link, and both sides of the coin are clearly worth addressing. After seeing the data, Dr. Dasgupta and colleagues formed the Virtual Patient Network (VPN) for peer support and action in youth and young adults with type 1 diabetes, a community of support with peer leaders. Mr. Jordan McCaron, one of the peer leaders who lives with type 1 diabetes, said that a lot of the activity in the VPN support group is simply sharing entertaining memes on the topic of daily life with diabetes that all members can relate to and laugh at. For example, a picture of Spongebob Squarepants falling backwards while crossing a doorway with the caption “when your pump tubing gets caught on the doorknob.” We would be interested to reassess the participants’ perceptions of stigma after participating in the VPN Facebook group: Does knowing that you’re not alone reduce feelings of shame?
Dr. Michael Vallis: Technology Can Bridge the Gap for Sustained Behavior Change
In a packed-to-capacity LifeScan symposium, Behaviour Change Institute’s Dr. Michael Vallis (a clinical psychologist focusing on diabetes/obesity care) advocated for HCPs to embrace collaboration with technology in order to empower patients. At the core of Dr. Vallis’ remarks was the importance of psychosocial aspects of diabetes care, aspects that are too often neglected or poorly addressed in interactions. To the room full of HCPs, Dr. Vallis remarked that “it’s not your job to judge behavior – but it is your job to understand behavior.” We loved to hear this sentiment, but how does an HCP go about understanding behavior? Dr. Vallis explained that HCPs must attempt to engage their patients in change-based relationships. These relationships are built on: (i) understanding the holistic patient experience; (ii) trying to uncover underlying personal reasons the patient may harbor in resisting change; (iii) asking permission and negotiating recommendations and educational requests (“you are not entitled to educate or give recommendations!” he explained); and (iv) recognizing that the patients themselves have the ultimate responsibility for their behavior; however, clinicians can empower patients to make better choices. This discussion of better negotiating change with patients harkened us back to Dr. Vallis’ remarks at Diabetes Canada 2017, where he similarly discussed how HCPs can help patients overcome common psychosocial barriers in treatment. Where does technology fit into this discussion? Dr. Vallis explained that a huge disconnect currently exists between HCPs understanding the importance of these psychosocial aspects of care and their competence in actually addressing them. Outlining a very broad vision, Dr. Vallis pointed towards technology as the solution for bridging this gap: SMBG apps that collect and interpret data and identify issues for patients and clinicians can be critical in HCPs better understanding their patients’ experiences and empowering patients in self-management systems. Furthermore, the adoption of technology can also ensure enhanced sharing of data between patients and HCPs, further dissolving barriers between these two all-important stakeholders.
- “Make it emotional, not intellectual.” When asked how to approach patients that are using technology (such as CGM) but who aren’t responding well to the glycemic trends that they present, Dr. Vallis responded with this creed of “make it emotional, not intellectual.” It’s ever-important to make sure that “tech is not pushed onto the patient.” Dr. Vallis advised framing the interaction with this greater emotional context and asking the patient “if it’s okay to stop for a second so we can take some time to reflect on how you feel about this technology.” Simply asking the patient about their gut feeling regarding this new piece of technology (thumbs up or thumbs down?) can yield great dividends in understanding the fundamental concerns they have in adopting it.
The portion of Abbott’s booth dedicated to FreeStyle Libre advertised the Canadian launch of the FreeStyle LibreLink app, which enables scanning of FreeStyle Libre sensors with a smartphone, allowing users to leave the reader at home. LibreLink has the added value of allowing for remote monitoring via LibreLinkUp. The app began rolling out in Canada on September 26, and though few people are aware of it at this time, a rep we spoke to said it was a no-brainer: “People get so tired of having to carry something extra around.” In an eight-second demo, she showed how easy it is to swipe the device, get a reading, and add contextual information related to diet and activity. Reps cautioned that the launch in Canada is currently limited to iOS, but Android is approved and will come soon. FreeStyle LibreLink has been available for both Android and iOS users in at least 12 European countries since February, though a previous version developed in partnership with AirStrip was previously available on Android phones. There is still no word on US timing for LibreLink, though we have to imagine that it will come sooner rather than later.
The Ascensia booth showed off the Contour Next One meter with the typical taglines “tap into brighter self-management” and “their diabetes illuminated.” Reps underscored their goal of making connected devices the norm – an uphill battle for sure, but fortunately one that much of the diabetes technology industry is fighting together. The new-and-improved Contour Diabetes app and its ability to identify glycemic trends and provide contextual prompts was also center-stage, and though it is early days, reps noted very positive feedback from attendees, particularly around the prompts and structured SMBG plans.
The Ultra-Fine Nano Pro 4mm pen needle, now available in Canada, was centrally displayed in the exhibit. The new needle apparently has a number of benefits: (i) an easier-to-grip outer cover; (ii) an easier-to-remove inner shield; and (iii) a redesigned “needle base” (see picture below) that is contoured to give a more reliable 4mm injection, per consensus recommendations. A rep told us that the needle is currently sitting at FDA and can be reasonably expected to enter the US market next year.
DarioHealth showed off its BGM, as well as the new Dario Engage platform (pictures below). The B2B2C offering allows providers to track and communicate with the most at-risk patients via phone call, email, and in-app chat. In the Coach Dashboard hub, providers can see a list of their patients, previous interactions, and risk-stratification tools such as Alerts (recent out-of-range readings) and “Patients to Engage” (those who haven’t logged blood glucose/interacted with HCP recently). Though the demand on the coach is fairly high today, DarioHealth will look to ease that burden; for example, a future feature will allow coaches to preset education plans for patients based on their characteristics, such that the appropriate content will flow on a set schedule without the provider having to curate or send it. In addition, they are in talks to bring remote coaches into the mix, should clinics prefer that their in-house providers spend more time delivering face-to-face care to more patients. Reps told us that the company is currently using a “pull-through” model, selling the platform to clinics at a very low price in hopes that will lead to greater numbers of patients buying meters and strips.
There was a slight aura of exasperation about the Dexcom booth, as the company eagerly await nods from Health Canada on the Clarity mobile app and, most of all, the G6. Unsurprisingly, educators stopping by the booth almost invariably led with, “when is G6 coming?” It sounds to us like G6 could come any day now – we heard rumors that the company had been prepared with G6-oriented panels, but they didn’t get to flash them at Diabetes Canada. In the meantime, reps detailed the G5 Mobile CGM system with the tagline “Know before you’re low.” When we saw this marketing from across the hall, we immediately thought that G6 must’ve been approved, and the line was a reference to predictive alerts. Of course, we were wrong, as it was a reference to G5’s trend arrows and alarms. When asked, reps told us that the slogan tested extremely well in Canada, whereas others pertaining to fingersticks didn’t elicit the same positive response. New to us, apparently “prick” is a very bad word to use in Canada, so “no more pricking your finger” to advertise an insulin dosing claim was out of the question!
The next-gen OmniPod Dash system is currently with Health Canada, but we learned that a piece of the Insulet booth veiled by curtains on all sides housed Dash, ready for attendee comment. Limited US market release of the new Dash touchscreen PDM commenced last month, with a full release still in early 2019, and we can only imagine Canadian OmniPod users are eager to access the improved user experience, phone app, remote follow capability, and more. A friendly rep at the booth was eager to tell us her story: She had worn pumps for 21 years and after getting the OmniPod recently, she decided that she simply had to work at Insulet. She loves that she doesn’t have to pull out her pump to bolus – and looks forward to a day when she will be able to do so from her phone/watch – and loves that she doesn’t have to manage pump tubing when she goes to the bathroom, walks past a doorknob, and in other trivial situations. We were also drawn to a form in the booth titled “Considering an Insulin Pump?”, which walked potential MDI-pump converts through the decision, including questions about hygiene, work, sleep, clothing and activity.
“Introducing 670G” was all the banner needed to say – attendees knew all about it and were excited to learn more. The MiniMed 670G system was approved by Health Canada earlier in the week, and Drs. Jen Sherr and Grazia Aleppo presided over a packed-to-the-brim symposium on the system and CareLink reports two days later. The first units are expected to begin shipping in Canada in December. Curiously, Medtronic continues to avoid using the term “hybrid closed loop” outside of the US, referring to 670G as “the world’s first self-adjusting insulin pump.” We’re not sure what went into the decision to alter marketing, but we’re not sure that “self-adjusting insulin pump” conveys any more clearly than hybrid closed loop that the pump only modulates basal rates and requires user interaction – perhaps the decision was more related to lexicon, as “hybrid closed loop” is far more technical. As a reminder, the standalone Guardian Connect Mobile CGM system was approved in May, but it is expected to launch in Canada in the fall. A rep told us that the MiniMed Mio Advance all-in-one disposable infusion set inserter launched in May to great excitement and was already out of stock by the summer, despite its slightly higher price point ($2.35/box vs. $2.15/box for other sets). We’ve found the inserter very intuitive, and necessarily so – a rep who formerly worked at Animas (and joked, “I lost my pump and my job”) told us that the #1 complaint to Animas was about difficulties with the Mio infusion set.
- Medtronic also had a tech suite showing attendees (under NDA) items in the pipeline that are set to launch up to five years out! We didn’t have the opportunity to see into the future, but we imagine most of what was shown can be found in the June Analyst Meeting slide deck.
The OneTouch Reflect BGM we first saw in the EASD exhibit hall is with FDA and Health Canada, and a rep estimated that it could launch in both geographies in the next 4-6 months. A CE mark is also pending, with launches in France and Germany expected “very soon.” The unique meter includes a “Blood Sugar Mentor” that provides on-meter patterns, insights, and encouragement to the user. Reflect will also send data to the OneTouch Reveal app, which is up for an update later this month. Currently in beta, the updates will bring patients push notifications when the app notices a trend (say, of hypoglycemia at a certain time of day) and an “A1c Comparator,” which as the name implies, compares A1c to an estimated A1c based on the 90-day mean glucose. This feature is much alike the “Glucose Management Indicator” (GMI; formerly eA1c) about to be re-integrated into CGM reports, as it is intended to help patients understand “where they run in relation to A1c.” Nice! We quite like the name “A1c comparator,” actually, since inherent to the term “compare” is the implication that it will likely be different.
Roche’s booth touted the Accu-Chek guide – emphasis on accuracy and spill-proof vial – and mySugr for the first time in Canada. Reps said there are ~80,000 mySugr users in Canada already, but this is the first time they’ve detailed the BGM integration or the app generally at a conference. Today, purchasers of Accu-Chek Guide have an automatic upgrade to mySugr Pro (premium version of the logbook app), but there is no coaching bundle in Canada just yet (reps believe there will be, eventually). Educators at the booth at the same time as we were are particularly excited about the passive logging, how CGM, BGM, and activity data all flow seamlessly into the app.
On Friday afternoon, Tandem heard that they had received Health Canada licensing for the t:slim X2 with G5 integration. This was not only a license to sell the product in Canadian provinces but also to take down the flimsy sign at the booth reading “pending approval” and talk about the pump to curious conference-goers for one day. The booth was notably more hopping on Saturday than any of the preceding days, and it added an even greater sense of excitement to a meeting that had just seen 670G approved two days before. According to the press release, Tandem plans to launch the pump on a province-by-province basis throughout 4Q18. It will be interesting to watch adoption dynamics of Tandem vs. Medtronic in this geography in the coming months, particularly after we hear more about Tandem’s plans to bring Basal-IQ (PLGS with G6) to Canada; 670G is expected to begin shipping in December, and Basal-IQ is set to roll out internationally in 1H19, though Tandem has never provided specifics on country cadence. From what we picked up at this meeting, 670G and the promise of closed loop (the “hybrid” part hasn’t fully sunk in, as was the case in the US) is all the rage among Canadian CDEs now, but there is a lot of time for Tandem to pick up marketing efforts before December. We imagine both products will have very successful rollouts and updates on their Q1 calls. Tandem’s Canada sales team is now ~10 people-strong, “rearing up and ready to go.” Management has previously said that they will operate through a logistical distributor in the region, but will rely on its own sales force; with ~300,000 type 1s and strong reimbursement, Canada is a big near-term market for the company.
AZ’s booth was dedicated to its SGLT-2 inhibitor Farxiga (dapagliflozin). Highlighting the exhibit was an interactive standing panel, where visitors could learn more about patient profiles that are well suited to Farxiga use. Large banners touted Farxiga as a second-line option after metformin – we’re glad to see this focus on early treatment intensification with these effective therapies. Farxiga is central to AZ’s diabetes business, having contributed over 90% of diabetes revenue growth in 2Q18 – and continues to show promising growth potential as more patients and providers utilize it early on in treatment algorithms.
To be sure, anticipation is building around the presentation of full results from DECLARE, the CVOT for dapagliflozin, at AHA 2018 this November. AZ’s medical representative touted the positive topline results from DECLARE, with dapagliflozin meeting one of its co-primary endpoints for reducing risk for hospitalization for heart failure and cardiovascular death over placebo, and the potential for dapagliflozin to garner a broader indication than its competitors in the SGLT-2 inhibitor class because of the larger primary prevention cohort used in the study.
Devoted to its SGLT-2 inhibitor Invokana (canagliflozin), the J&J booth highlighted the drug’s mechanism of action and potency in information-rich handouts and panels. In a fruitful conversation with the medical information representative at the booth, we learned that CREDENCE data was just locked a few weeks ago (although the announcement of CREDENCE being stopped early for meeting its primary endpoint on renal protection was made in July, it has taken a few months to complete final follow-up patient visits and the like). Moreover, the expectation is for full CREDENCE results to be released sometime “around Christmastime” – most likely in early January. We surely can’t wait for these results! As a reminder, CREDENCE’s primary endpoint (which was met early, thus the early stop date) was a composite of end-stage kidney disease, doubling of serum creatinine, and renal or CV death – it will be key to see if any of these individual factors drove the positive results, or if benefits were seen across the board. Nonetheless, we certainly expect J&J to eagerly submit this data for a potential label update, which would greatly enhance Invokana’s ability to differentiate itself amongst a crowded and competitive SGLT-2 inhibitor class. Positive CREDENCE news definitely comes at a critical junction for Invokana, as it has struggled recently, posting a 27% YOY decline in 2Q18 (however, we do note that sales have been stronger outside of the US, where sales increased 18% YOY in 2Q18). Management has blamed this relatively weak financial performance on increasing discounts, higher rebates, and competition in the class; we would imagine that the latter reason could be directly addressed with these positive CREDENCE results and serve as a strong tailwind for the Invokana franchise.
Lilly’s booth had a cozy feel and was highlighted by the opportunity for passersby to pick the next color that an artist used to paint a scene on a canvas. Theatrics aside, Lilly’s exhibit focused on Basaglar (biosimilar insulin glargine) and Trulicity (dulaglutide), with a specific emphasis on Trulicity’s ease of use. Large signs touted Trulicity’s simplicity: once-weekly dosing and simple design of its auto-injector pen (“Just uncap, place and lock, and then press and hold.”) Trulicity has seen significant growth recently (+62% YOY growth to $780 million in 2Q18) and is now neck-and-neck with competitor Victoza (liraglutide) in terms of total prescription volume. Results from the highly-anticipated REWIND CVOT will play a role in determining whether Trulicity can maintain this strong trend (or even accelerate). Management has revealed that topline results for the CVOT are expected in 4Q18, and we’re on the edge of our seats. REWIND notably enrolled a very large primary prevention cohort (69%), giving Trulicity the potential to garner a much broader CV indication if it does show CV benefit in the trial. Such a label would be an undoubted win, especially in the context of the ever-competitive GLP-1 class that constantly demands innovation and progress to remain relevant. REWIND discussion aside, we’re surely glad to see Lilly maintain a presence in the exhibit hall – we were quite surprised by their lack of a full-fledged booth at AADE 2018.
The Danish diabetes giant boasted one of the larger booths in the somewhat cramped exhibit hall, highlighting Ozempic (semaglutide), Fiasp (rapid-acting insulin aspart), Saxenda (liraglutide 3.0 mg), and Tresiba (insulin degludec). As we’ve come to expect at Novo Nordisk’s booths recently (see EASD 2018; ADA 2018; AADE 2018), Ozempic took center stage, with much of the presentation focused on the new injectable semaglutide which is still early in its launch cycle. This emphasis on Ozempic falls in line with management’s intention of switching over promotion to Ozempic from Victoza in most markets. Notably, banners and panels for Ozempic have lined the stairs and escalators at Diabetes Canada, promoting its two pens (one for the 1 mg dose and a separate pen for the 0.25 and 0.5 mg doses, similar to what’s marketed in the US. Notably, in Europe, there are three pens: one for each of the separate doses). [There was even Ozempic ad material greeting and bidding farewell to conferencegoers at the Halifax airport.] Tresiba was also featured at the booth and in the conference at large; it similarly obtained placement along stairs and escalators in the conference center. Specifically, posters touted the fact that Tresiba is now covered by multiple provincial formularies across Canada – namely Saskatchewan, Manitoba, Ontario, Quebec, and Newfoundland. Tresiba launched in Canada last October, and was notably the first insulin to be added to Canada’s national Register of Innovative Drugs. Seeing as Canada does not provide universal prescription drug coverage, we’re glad to see that more patients are starting to gain affordable access to Tresiba in multiple Canadian provinces.
Sanofi’s exhibit hall booth was ripe with information spanning its diverse diabetes portfolio. We were pleased to see a dedicated section for Soliqua, Sanofi’s fixed-ratio combination therapy of lixisenatide and insulin glargine. A large panel for Soliqua highlighted the slogan, “When A1c is high, act with Soliqua” and touted its efficacy with data from the LixiLan trial. Representatives at the booth noted that Soliqua just officially launched in Canadian markets a few weeks ago, and that access is still building as a result; however, they spoke positively on the fact that roughly 70% of plans in Canada now have access to the fixed-ratio combination therapy. Soliqua was also advertised for its effectiveness in particularly combatting post-prandial glucose increases, due to the shorter-acting nature of lixisenatide; this comes in direct comparison to the longer-acting nature of liraglutide, Novo Nordisk’s GLP-1 Victoza that is used in its fixed-ratio combination therapy, Xultophy (insulin degludec and liraglutide).
Sanofi also highlighted its R&D pipeline in the booth, headlined by sotagliflozin, the Lexicon-partnered SGLT-1/2 dual inhibitor currently being studied as an adjunct therapy in type 1 patients. A detailed panel emphasized that sotagliflozin has been submitted for approval in this population in both the US and EU; conversations with a medical representative revealed that it has also been put forward for approval with Health Canada in the past few weeks. Phase 3 EASE results, presented just recently at EASD 2018, have surely provided ample fodder to consider regarding SGLT inhibitor use in type 1 patients. The big question as always with this adjunct therapy is how to mitigate DKA risk; EASE results hinted at the potential of a lower dose of empagliflozin (2.5 mg) in minimizing DKA events (albeit with reduced glycemic efficacy). The phase 3 inTandem trials for sotagliflozin did not investigate any such lower doses that could have this potential, as DKA events persisted across the different doses used. We’ll be eager to see how regulatory agencies respond to sotagliflozin’s submission.
--by Martin Kurian, Brian Levine, and Kelly Close