March 11-13, 2015; London, UK; Full Report – Draft

Executive Highlights

The 2015 Diabetes UK Professional Conference, held in London from March 11-13, featured an exciting slate of presentations on everything from closed-loop data to CV safety of diabetes drugs to the launch of a new National Diabetes Prevention Programme. This year’s conference set a new attendance record of 3,300, and we thought the exhibit hall seemed bigger and better as well, perhaps reflecting the UK’s recent tendency to lead the Northern hemisphere in the introduction of drugs and devices (e.g. Novo Nordisk very recently made Xultophy (IDegLira) available in the UK, Medtronic made the 640G available here three weeks ago, and the UK is one of the few places that you can obtain Abbott’s Libre – if you are lucky). On a more sobering note, Diabetes UK suggests that there are currently 3.8 million people with diabetes in the UK, of which 630,000 are undiagnosed – clearly, the need for new therapies and screening programs could not be greater. See below for several of our top highlights from the meeting, followed by our more in-depth coverage.

1. Mr. Simon Stevens, the CEO of NHS England, launched the National Diabetes Prevention Programme (modeled after the milestone DPP study in the US), saying that “obesity is the new smoking.” Mr. Stevens commented, “if this was a pill, we would be popping it.”

2. Dr. Roman Hovorka (University of Cambridge, Cambridge, UK) presented an overview of his thousands of days and nights of closed loop experience to date.. Overnight results are stellar, and the daytime is definitely an improvement over open loop, but less spectacularly so because of the meal challenges. His future concerns have less to do with the science and the devices and relate more to human factors, patient selection, commercialization and reimbursement.

3. Dr. Rury Holman (The Oxford Centre for Diabetes, Endocrinology and Metabolism, Oxford, UK) presented a predictive model showing that a 1% drop in A1c results in a 14% lower relative risk of a heart attack.

4. Dr. Harry Hemingway (Farr Institute, London, UK) discussed the need to get serious about incorporating ‘big data’ into the National Health Service and showed spectacular data on the cardiovascular risks of type 2 diabetes and the association between higher A1c and twelve sub-types of heart disease.

5. Dr. Nick Finer (University College London, London, UK) made the case that the NHS should gear up to do thousands more bariatric surgeries each year, pointing out that the procedure literally saves lives (mortality is halved after the post operative period) and actually makes money for the health service after only three to four years.

Honorable Mention: Always media-savvy, Diabetes UK was proudly displaying the Daily Express headlines that screamed “Vaccine Will Cure Diabetes. ” This was in response to the T1DUK Immunotherapy Consortium, a new £4.4 million investment in type 1 diabetes immunotherapy research sponsored by Diabetes UK, Tesco (a retailer) and JDRF, and modeled somewhat on TrialNet (which also exists in the UK).

Detailed Discussion and Commentary

Hot Topics I

Contemporary Vascular Risks In Type 2 Diabetes

Harry Hemingway, MB.ChB (Farr Institute, London, UK)

This talk by Dr. Harry Hemingway discussed the significant opportunities the UK has to further exploit ‘big data’ from the NHS for higher-resolution epidemiology and research across the translation spectrum. Britain is unique in that every citizen is registered with a doctor, everyone uses the same electronic health records, everyone has a unique health ID, and there is a wide range of linkage possibilities. In theory, it should be possible to do cohort studies on 50 million people. However, converting the raw data into ‘research ready’ data is not trivial for many reasons, and other countries (notably Sweden and Denmark) have made more visible progress. The Farr Institute has been established to get more value from this data. As an example, Dr. Hemingway shared data showing elevated hazard ratios for ten out of twelve sub-types of cardiovascular disease in patients with type 2 diabetes. He noted that CV composite endpoints in many trials may not reflect the highest risk sub-types, as a means of demonstrating the value of bringing big data to bear on future trial design. Finally, he showed that there is a strong association between A1c and cardiovascular risk for at least eight of the twelve sub-types of cardiovascular disease – as Dr. Rury Holman outlined in his talk in the same session (see below), this finding is consistent with results from several long-term randomized trials .

• The NHS is unique in having access to very large cohorts, allowing high phenotypic resolution. In the UK, every citizen is registered with a doctor, everyone uses the same electronic health records, everyone has a unique health ID, and there is a wide range of linkage possibilities. Having said that, converting raw data to research ready data is not trivial. CALIBER is an online portal developed by the Farr Institute that performs this role. Dr. Hemingway gave examples of pooled and analyzed blood pressure outcome data and they were tremendous – they included huge numbers of patients, had great statistics, and looked really clean.
• Dr. Hemingway shared data showing elevated hazard ratios for most sub-types of cardiovascular disease in patients with type 2. Strangely, the risks for subarachnoid hemorrhage and abdominal aortic aneurysm were lower in type 2 diabetes (both hazard ratios are around 50%), which is apparently consistent with previous knowledge. For the other ten sub-types, patients with type 2 diabetes had elevated hazard ratios (e.g. 2.98 for peripheral arterial disease, 1.78 for ischemic stroke, 1.56 for heart failure, 1.54 for nonfatal MI).  The full list of twelve sub-types was: stable angina, unstable angina, non-fatal MI, unheralded coronary death, heart failure, arrhythmia or sudden cardiac death, transient ischemic attack, ischemic stroke, subarachnoid hemorrhage, intra cerebral hemorrhage, peripheral arterial disease, abdominal aortic aneurysm.
• Data also shows that CV hazard ratios increase with increasing A1c in at least eight out of the twelve sub-types. The data also shows that there is not a particularly strong protective effect for women versus men, a common misconception.
• Dr. Hemingway presented lifetime risk curves, which emphasized the much greater risks of angina, MI and peripheral artery disease in type 2 diabetes. Note that these are not typically included in the primary composite endpoint of clinical trials, which tend to focus on nonfatal MI, nonfatal stroke, and CV death. We have also heard a number of researchers suggest that heart failure should be a component of the primary endpoint – EXAMINE investigator Dr. Faiez Zannad (Institut Lorrain due Coeur et des Vaisseaux, Vandoeuvre-lès-Nancy, France) made this point recently in an interview with us, suggesting that heart failure is a greater clinical concern in patients with type 2 diabetes than stroke or MI. Perhaps observational data like this can be used to refine clinical trial design in the future.

Intensive Glucose Lowering And Macrovascular Risk

Rury Holman, FMedSci (The Oxford Centre for Diabetes, Endocrinology and Metabolism, Oxford, UK)

The inimitable Dr. Rury Holman (referred to as a ‘young man’ by the chairman) aimed to quantify the link between glycemic control and risk reduction of cardiovascular (CV) disease. His valuable analysis suggested that over time, a 1% drop in A1c is associated with a 14% drop in the risk of myocardial infarction in type 2 diabetes. A number of large-scale trials confirm this result beautifully. He also noted that in UKPDS, A1c was an independent risk factor for coronary artery disease. Dr. Holman then turned his attention to ACCORD, ADVANCE and VADT, trials that largely fell short of expectations by failing to show a significant reduction in macrovascular complications with intensive glycemic control (we personally think the trials may not have been long enough). However, a meta-analysis of these trials and UKPDS showed that a 1% A1c reduction was associated with a 15% relative risk reduction for CVD. In the case of ACCORD, he argued that the finding of increased mortality in the intensive group doesn’t necessarily show that a low A1c is bad, just the means for achieving it. We would speculate that a comparable study that achieved intensive control with agents other than insulin and sulfonylureas would have a greater chance of demonstrating benefit – we would love to see such trials in the future, though we realize many investigators may not see them as worth the high cost in time and resources. We look to others, perhaps governments and philanthropists, to fund these, given the major scope of the diabetes epidemic.

• People with diabetes have double the risk of heart disease. However, the correlation with exposure to glucose over time is not nearly as strong for MI as it is for microvascular disease.
• A landmark analysis of the UKPDS results predicted that a 1% drop in A1c is associated with a 14% drop in the risk of MI. The analysis showed that average A1c over time was linked to risk of MI, but the association wasn’t as strong as with microvascular disease. Another meta-analysis of six type 2 CV trials showed an overall risk reduction of 19% with intensive therapy. In type 1 diabetes, the relative risk reduction was 62% in a meta-analysis, and 42% in DCCT/EDIC.
• A meta-analysis of the ACCORD, ADVANCE, UKPDS and VADT trials showed a 15% relative risk reduction for CVD with a 1% reduction in A1c. The four trials were the major outcome trials investigating the link between glycemic control and cardiovascular disease; none was able to demonstrate a significant risk reduction with intensive vs. conventional control. Dr. Holman argued that the finding of increased mortality with intensive control in ACCORD didn’t show that low A1cs were bad, just the means used to achieve a low value. Dr. Howard Weintraub (New York University, New York, NY) offered similar sentiments recently at ACC, noting that the trial showed a trend toward improvement in the primary composite outcome, a lower mortality risk in lower A1c categories, and a reduction in non-fatal MI with intensive control.
• The PROActive trial demonstrated a 16% risk reduction in a secondary MACE composite endpoint with Takeda’s Actos (pioglitazone), but a simulation with the UKPDS Lifetime Simulation Outcomes model suggested that the benefits were fully explained by conventional risk factors, including the A1c lowering. TZDs, of course, come with a substantial amount of safety baggage, and the class has largely fallen out of favor in recent years. This is understandable, though we are under the impression low-dose TZDs may well be useful – they have also fallen enormously out of favor, however, even though all the negativity may be far lower with them. They do impact insulin resistance, and we’d like to see more study here. The “regular” dose TZDs are dead on arrival these days given the baggage and we agree it’s an impossible argument to make with patients to use these therapies that cause weight gain, for a start, along with the edema, association with CHF and bone issues, etc.
• The UKPDS model also shows that early glycemic intervention significantly lowers the risk of death. In a simulation of a patient diagnosed at age 50, waiting ten years before a 10 year period of A1c controlled at 7% reduced the chance of death by 7%. If the patient was controlled at an A1c of 7% from diagnosis, then risk of death is reduced by 19%.
• In people with diabetes, cardiovascular disease is the most common cause of death. There is a high residual risk, even after treating typical common risk factors, so improving glucose control offers modest additional reductions in risk. The next generation UKPDS risk engine will contain a ‘what if’ tool to model the impact of adding various drug therapies.

Hot Topics 2

Safety Of Glucose Lowering Therapies: An Update

Stephen Gough, MD (Oxford Centre for Diabetes Endocrinology and Metabolism, Oxford, UK)

Dr. Gough had the task of summarizing the latest safety data or safety concerns for each of the major drug classes for type 2 diabetes. His bottom line was that there are no new major safety concerns. But he started by lambasting the NICE draft guidance on type 2 diabetes, which suggests that metformin and glinides should be used before any more expensive drugs. He felt that was unsafe for patients. He argued that “we should feel a bit more confident with TZDs” since the risk of bladder cancer and cardiovascular disease has faded. SGLT-2 inhibitors have a higher ‘real-world’ drop out rate, primarily because of genital mycotic infections, but they are not too serious, and other adverse effects seem minor (though there have been recent anecdotal reports of euglycemic DKA – see below). On DPP-4 inhibitors, he believes we have laid to rest any concern about pancreatitis and pancreatic cancer, and they have no relative risk of cardiovascular disease based on the CVOTs (SAVOR and EXAMINE) that have reported thus far. Many major cardiovascular outcomes studies are set to report over the next five years (including TECOS for Merck’s Januvia [sitagliptin] and ELIXA for Sanofi’s Lyxumia [lixisenatide] at ADA in June), so we will get to see some rich safety data.

• The current NICE draft guidance for type 2 diabetes deserves a ‘major health warning’ according to Dr. Gough, who believes that the recommendations are ‘unsafe’. The new guidance is scheduled to come into effect this summer, and comments have just closed. The new guidance will recommend that physicians use insulin secretogogues (e.g. replaglinide and sulfonylureas) before any of the newer therapies are employed – presumably for cost reasons. While we recognize that generic treatments may be the only option in some cases due to cost, we agree with Dr. Gough that the exclusion of newer therapies from the guidance is not consistent with the latest science – overall, we think this is really an unfortunate turn.
• Dr. Gough believes that SGLT-2 inhibitors, DPP-4 inhibitors, and GLP-1 agonist/insulin combinations should have much greater prominence in the guidelines, as they do in the ADA/EASD guidance.  He is a fan of the ADA/EASD guidance because it also lists the advantages/disadvantages of the various therapies.
• In Dr. Gough’s view, TZDs have been somewhat redeemed over recent years, although they have well recognized side effects (weight gain, peripheral edema, bone fractures). We should be getting ten-year safety data soon for Takeda’s Actos (pioglitazone). However, bladder cancer concerns seem fully allayed for both Actos and GSK’s Avandia (rosiglitazone), and the FDA removed restrictions on rosiglitazone related to cardiovascular disease at the end of 2013. That said, we believe the class’ negative reputation will be difficult to shake off, as sales of both Actos and Avandia have been on a downward trend for years stemming from negativity over safety and side effect profile and Actos is currently the subject of litigation involving nearly 9,000 patients in the US.
• The main side effects of SGLT-2 inhibitors are genital mycotic infections, which affect 11% of women and 4% of men, according to Dr. Gough. While we have often heard speakers downplay the severity of these effects, it appears that clinical trial data may underestimate their significance: Dr. Gough presented a ‘real-world’ review suggesting that 22% of patients discontinued dapagliflozin because of side effects, compared to 3-4% in clinical trials. Leaving aside genital infections, other adverse effects such as urinary tract infections, volume depletion, urinary calcium excretion, and small increases in LDL cholesterol are not particularly significant at this stage. However, Dr. Gough also noted that caution is warranted regarding the use of SGLT-2 inhibitors in type 1 diabetes. While reports are largely anecdotal, there does appear to be an elevated risk of euglycemic DKA, exacerbated by insulin omission by the patient.
• DPP-4 inhibitors are undergoing many cardiovascular safety studies. The ones that have reported to date have shown absolutely no difference in hazard from placebo. There also appears to be no risk of pancreatitis and pancreatic cancer and the FDA and EMA have said so in writing. There is a mysterious connection to an excess of hospitalization for heart failure, but the jury is out at present. The next big study, TECOS, might help shed some light, and it’s due to report at ADA 2015. The FDA is holding an EMDAC meeting on April 14^th to discuss the results of the SAVOR and EXAMINE trials (the two DPP-4 inhibitor CVOTs that have reported thus far), and we assume that heart failure will be a central topic. We continue to wonder why this was scheduled prior to TECOS findings, which will be at ADA on Monday afternoon at the Scientific Sessions.
• The EMA recently issued a positive opinion on the use of Novo Nordisk’s Victoza (liraglutide) in moderate renal impairment. Formerly they had not recommended it. In studies, there was no change in renal function from baseline, but an improvement in A1c and weight.

Type 2 Diabetes Therapeutics

DPP-4 Inhibitors, GLP-1 Receptor Agonists And The Cardiovascular System

Anthony Barnett, MD (University of Birmingham, Birmingham, UK)

There is an epidemiological link between glucose and cardiovascular (CV) risk, but whether improving glycemic control improves CV outcomes and mortality is controversial. There is evidence that long-term glycemic control leads to a CV benefit, but there doesn’t seem to be a benefit for older people or those with a longer duration of diabetes who haven’t formerly been in good control – in fact, intensive control has been negative in some trials in this group. There are now many ongoing CV outcomes trials with agents that don’t lead to hypoglycemia or weight gain (possibly implicated in negative outcomes of intensive control) which have reported results (SAVOR, EXAMINE, and ELIXA) or will do so over the next few years. However, the consensus seems to be that these trials are unlikely to uncover any CV benefit that exists, as they are designed as safety studies and typically have short durations and enroll very high-risk populations.

• Dr. Barnett reminded the audience of the rosiglitazone (GSK’s Avandia) saga that led to the requirement of large-scale cardiovascular outcomes trials for new diabetes drugs. As a reminder, rosiglitazone received enormous publicity for a claimed association with cardiovascular (CV) disease, which the FDA belatedly concluded was less significant than originally claimed. As an interesting aside, Dr. Barnett suggested that sulfonylureas share many similar issues, but haven’t been sanctioned because they haven’t had the publicity and are very cheap. He showed some all-cause mortality trial data he obtained that was suggestive of an increased risk with sulfonylureas (this has been widely spoken about for years though not necessarily proven – FDA though as we understand it doesn’t even allow this class in the major outcomes trials!), and listed many potential issues including ischemia, endothelial dysfunction and inflammation.
• There are many short-term studies suggesting that incretin-based therapies improve cardiovascular risk factors. These are supported by a meta-analysis of pooled phase 2 and 3 DPP-4 inhibitor trials, which show a trend toward cardiovascular safety. However, the total number of events is low, and there are many limitations with the study design and protocols that mean we should proceed with caution.
• Four major GLP-1 agonist and five major DPP-4 inhibitor CV outcomes trials are reporting through 2019, which will yield fascinating information. It is important to note that these are safety trials that are designed to demonstrate non-inferiority for regulatory purposes and not necessarily intended to prove cardiovascular benefit. The study populations are likely too old and too ill to show any cardiovascular protection from A1c reduction, and in any case, A1cs are going to be similar in the various arms. See the table below for a list of key trials and their expected completion dates.

Table: Status of Current CVOTs for Incretin Therapies

• Both EXAMINE and SAVOR-TIMI (which was huge at n=16,492) showed absolutely no difference between alogliptin and saxagliptin (respectively) and placebo in terms of the primary composite CV endpoint. Superiority was not shown, but these were relatively short studies. In admittedly different populations, STENO-2 took eight years for the curves to diverge and UKPDS needed 20 years to demonstrate statistically significant differences in macrovascular outcomes. Topline results from ELIXA announced earlier this month also demonstrated CV non-inferiority, but not superiority, with lixisenatide vs. placebo.
• Neither drug showed any evidence whatsoever of pancreatitis and pancreatic cancer, hopefully laying the issue to rest.
• In SAVOR, saxagliptin had statistically significantly increased rates of hospitalization for heart failure (HF) but no increase in related mortality, for unknown reasons. It could be a real effect or a chance finding. Alogliptin did not show a significant increase in HF hospitalization in EXAMINE, though we have heard debate over the implications of the non-significant 19% increase shown in a post-hoc analysis. The much-awaited TECOS, a similar large study of sitagliptin that will report at ADA 2015, will be important in clarifying the issue.
• The upcoming slate of GLP-1 agonist studies may reveal cardioprotective or reno-protective effects. Many of the caveats in the trials that have reported results (relatively short duration and high-risk populations) will apply to these trials as well, and we expect that neutrality is the most likely result. However, we did hear surprisingly optimistic commentary from Novo Nordisk recently on the potential for LEADER to demonstrate cardioprotection. Management suggested that the greater exposure with a long-acting GLP-1 agonist like Victoza compared to a shorter-acting agent or a DPP-4 inhibitor makes a superiority finding more likely (this logic would also apply to Trulicity). We can’t wait to hear more on this!

Insulin Combinations

Stephen Gough, MD (Oxford Centre for Diabetes, Endocrinology and Metabolism, Oxford, UK)

Most patients with type 2 diabetes on insulin are not reaching glycemic targets and adequate insulin dose titration is limited by hypoglycemia and weight gain. Dr. Gough presented evidence suggesting that combination therapy with insulin and newer agents (DPP-4 inhibitors, GLP-1 agonists, and SGLT-2 inhibitors) may provide an increased opportunity to get more patients to target. GLP-1 agonist/basal insulin combinations in particular have produced very impressive results in clinical trials (A1c reductions close to 2% with less weight gain than insulin and less nausea than GLP-1 agonists). Dr. Daniel Drucker (Lunenfeld-Tanenbaum Research Institute, Toronto, Canada) speculated recently at Rachmiel Levine that within ten to fifteen years, the vast majority of patients with type 2 diabetes needing injectable therapy will be treated with a GLP-1 agonist/basal insulin combination, which he described as the most effective way to treat type 2 diabetes, “bar none.”  

• Insulin therapy is limited by hypoglycemia and weight gain. As Dr. Rury Holman argued in his talk on glucose lowering and macrovascular risk, the finding in ACCORD of increased mortality in the intensive group does not necessarily imply that good glycemic control is dangerous, just that trying to get there too quickly (particularly with drugs that carry a risk of hypoglycemia) could be dangerous for some people. The key takeaway is that we need individualized therapy. In the last ten years, DPP-4 inhibitors, GLP-1 agonists, and SGLT-2 inhibitors have been approved and have all been studied alongside insulin, opening up new opportunities for a more targeted approach.
• Adding DPP-4 inhibitors to insulin can potentially improve A1c (by 0.5%-0.8%) without weight gain or risk of added hypoglycemia (assuming the insulin dose can be reduced).
• Adding GLP-1 agonists to insulin leads to improvements in A1c with no additional hypoglycemia and no weight gain. An insulin-sparing approach will have a modest glycemic effect but produce significant weight loss. Novo Nordisk’s Xultophy (insulin degludec/liraglutide) is now available for prescription in the UK. Results from Xultophy in treat to target trials showed that the combination was superior to either component alone (A1c reductions of 1.8% vs. 1.4% with insulin degludec and 1.2% with liraglutide after 52 weeks in the DUAL I phase 3 trial), and typical GI side effects from GLP-1 were reduced compared to the GLP-1 agonist alone. Sanofi’s LixiLan (lixisenatide/insulin glargine) is currently in phase 3, with results expected in 3Q15. In treat to target trials it showed a modest but significant A1c improvement vs. Sanofi’s Lantus (insulin glargine) with 84% of the LixiLan group achieving an A1c less than 7%, compared to 78% with glargine. LixiLan also produced better postprandial control and offered a weight benefit. Sanofi plans to submit LixiLan to regulatory authorities in 4Q15 in both the US and the EU.
• Adding SGLT-2 inhibitors to insulin delivers A1c reductions of 0.7%-0.8% with minimal impact on hypoglycemia (as expected) and weight loss (1-3 kg). In studies, there was no increase in insulin dose over 52 weeks in the SGLT-2 inhibitor arm, while the dose in the placebo arm increased consistently with the progression of the disease.

SGLT-2 Inhibitors: Recent Trials And Clinical Experience

John Wilding, DM (University of Liverpool & Aintree Hospital, Liverpool, UK)

This review of SGLT-2 inhibitors produced an impressive flow of slides. The three licensed SGLT-2’s in the UK (AZ’s Forxiga [dapagliflozin], Lilly/BI’s Jardiance [empagliflozin], and J&J’s Invokana [canagliflozin]) have been extensively studied. They deliver A1c reductions of 0.7-1.0% and 1.5-4 kg weight loss across a wide range of situations and combinations, including as adjuncts to insulin. They seem to prevent dose increases of insulin or worsening of kidney function over the long term, and there is evidence that they are cardioprotective, which is being investigated in many large long-term trials.

• SGLT-2 inhibitors reduce renal glucose reabsorption, eliminating an equivalent of approximately 300 calories/ day of glucose. Based on this mechanism, they would be expected to deliver a reduction in A1c, weight loss and a reduction in blood pressure.
• The three commercially available SGLT-2 inhibitors in the UK have been extensively tested, and are still undergoing clinical trials. These are AZ’s Forxiga (dapagliflozin), Lilly/BI’s Jardiance (empagliflozin), and J&J’s Invokana (canagliflozin).
• In clinical trials, SGLT-2 inhibitors deliver a 0.7-1.0% A1c reduction in almost any situation or combination. Dr. Wilding showed a plethora of trial data that demonstrated this convincingly. In head to head trials against other oral agents, Jardiance performed similarly to Merck’s Januvia (sitagliptin), and Forxiga demonstrated slightly better A1c results than a sulfonylurea after two years, but with a large improvement in weight. Invokana showed a slight improvement in A1c vs. Januvia at the high dose, again with much better weight loss, and with a larger percentage of patients achieving both A1c reduction and weight loss.
• SGLT-2 inhibitors can “break the cycle” of continually increasing insulin doses if they are added to insulin. In trials, Forxiga led to a placebo-adjusted reduction in A1c of about 0.5% (at the high dose) after two years when added to insulin. Weight was greatly improved (a 1.5% loss versus a 1.5% gain), but importantly over two years, insulin dose was flat for the higher doses of Forxiga, while the placebo arm increased consistently by 18 units/day over two years.
• All the SGLT-2 inhibitors deliver a consistent weight loss of 1.5-4 kg depending on the dose and trial situations. This is true in monotherapy, dual therapy, triple therapy and add-on to insulin.
• The drop in systolic blood pressure is typically 4-5 mmHg in clinical trials, compared to placebo. Patients starting with elevated blood pressure see a bigger effect (6-7 mmHg), which is enough to be clinically significant. Analysis shows that 40% of the reduction in blood pressure is due to weight loss and 60% is due to the effect of the drug.
• Typically, SGLT-2 inhibitors require good kidney function to be effective. Once eGFR drops below 60 ml/min/1.73 m², they are of less value. But there is evidence that long-term use of SGLT-2 inhibitors stabilizes renal function (i.e. eGFR stays flat). The ongoing CREDENCE renal outcomes study is investigating the potential protective effects of Invokana on diabetic nephropathy; a positive result would be an enormous win for the class.
• In the “real world” of primary care, outcomes are similar to the clinical trials. In a retrospective sample taken from a UK database of 40,000 patients on Forxiga, A1c reduction was 1.1% and weight loss was 6 kg at six months, although this does not include patients who discontinued. In another trial, a significant number (28%) of patients withdrew from therapy because of genital and urinary tract infections. Dr. Wilding suggested that if patients are counseled initially and they self-treat promptly, this side effect can be managed.
• Finally, the cardiovascular (CV) results in the regulatory trials show non-inferiority, but are suggestive of a positive effect. Many very large CV outcomes trials are underway and will conclude in the next five years. These include EMPA-REG OUTCOME (n=7,000; completion April 2015; results expected at EASD) for Jardiance, DECLARE for Forxiga (n=17,000; completion 2019), and CANVAS (n=4,365; completion 2018), CANVAS-R (includes renal effects; n=5,700; completion 2017), and CREDENCE (renal outcomes; n=3,700; completion 2020) for Invokana.

Opening Plenary Session

Introduction

The Rt Hon Baroness Barbara Young (Chief Executive, Diabetes UK, London, UK)

The inspirational Baroness Barbara Young, the president of Diabetes UK, introduced the conference with a call for more consistent care for diabetes in the UK. She announced a new England-wide National Diabetes Prevention Programme, which will be operational in two years. It will offer patients evidence-based programs of diet and exercise and will be funded by the National Health Service (NHS). Referrals to the program can come from physicians, risk assessments, pharmacies, workplace assessments, or self-referrals. The aim is to reduce suffering and reduce cost for the NHS. Additionally, Diabetes UK announced a new program to conduct trials of immunotherapies and vaccines for type 1 diabetes, analogous to TrialNet (which also exists in the UK and has screened 6,000 people). The program is funded (over £4.4 million/$6.5 million) by Diabetes UK, Tesco (a major retailer), and JDRF. In recent years, complication rates are down, but absolute complications are up, and incentive schemes driving delivery for doctors appear to have plateaued. So this summer, Diabetes UK will be working on a campaign to increase education for diabetes self-management.

Opening Plenary: The Future Management of Type 1 Diabetes Lies In….

Immunotherapy

Colin Dayan, PhD (Cardiff University School of Medicine, Wales, UK)

Dr. Dayan gave an exciting and optimistic presentation about the past, present and future of immunotherapies for type 1 diabetes. He noted that beta-cell preservation via immunotherapy offers the promise of device-free control even for less adherent patients, that several trials have shown temporary benefit at low toxicity, and that many more approaches are being developed. He announced a major strategic trials programme (the T1DUK Immunotherapy Consortium) that will be required to test agents in newly diagnosed patients and is expected to even allow prevention programs.

• Achieving ideal glycemic control with insulin alone is too difficult, as evidenced by the low percentage of patients at goal. Insulin therapy is limited by hypoglycemia and complications, so a little success in immunotherapy might go a really long way. Dr. Dayan showed a slide indicating that only 16% of patients had an A1c of 7% or below. 52% had an A1c of 8% or above, 24% were at 9% or above, and 3% were above 12%.
• We now know that there are over 50 genes implicated in type 1 diabetes, and the vast majority are connected to the immune system. These also work in combination and there is a large overlap with other auto-immune diseases.
• The number of trials in type 1 diabetes might appear large, but compared to other auto-immune diseases, type 1 has actually had fewer trials, and fewer successes. A GSK-sponsored trial of otelixizumab/ChAglyCD3 – an anti-CD3 therapy – found that a five-day treatment gave a benefit lasting for four years in some patients, but side effects included the reactivation of Epstein Barr virus (EBV). A number of anti-CD3 trials show that there exists a ‘sweet spot’ of C-peptide preservation but no re-activation of EBV, but it will take work to figure out the right products, dosing and protocols. Rituximab – an anti-CD20 therapy – demonstrated a benefit up to three months, but then C-peptide began to decline, though it still maintained a gap compared to the control group. A similar effect occurred with abatacept (which Dr. Dayan characterized as a low-toxicity intervention), with benefits going out to six months, a low adverse event rate and no EBV reactivation. Work is also ongoing in antigenspecific therapy, often referred to as diabetes vaccines. Additionally, the type 1 diabetes field is beginning to assess the biologics successfully developed to address other auto-immune diseases such as rheumatoid arthritis, psoriasis, Crohn’s disease and multiple sclerosis.
• At Cardiff University, Dr. Dayan has been taking part in a phase 1b study to establish the safety of a ‘diabetes vaccine’ – the MonoPepT1De Trial. At this point, he believes the treatment to be very safe. The vaccine is a peptide derived from proinsulin cells. Solid microneedles coated with the peptide can deliver the vaccine almost painlessly. Dr. Dayan is also developing new techniques of lymph node sampling in humans to more directly assess the effects on the immune system.
• Looking to the future, it’s expected that immunotherapies will be used in combinations. For example, drugs like rituximab and abatacept could be administered serially with a wait period of a few months in between. The goal would be to obtain a honeymoon period that lasts forever. But there are obstacles – a lack of adequately powered trials, a lack of biomarkers, finding protocol dose optimization, being able to replicate results, and utilizing combination therapy (often the companies don’t feel it’s in their best interests).
• Cardiff will be part of the new “T1DUK Immunotherapy Consortium,” sponsored by Tesco (a retailer who is donating over £3.3 million [~$4.9 million]) and JDRF, who is donating over £1 million (~$1.5 million). The group will create 15 centers across the country. Dr. Dayan feels that the UK has some advantages in developing a network similar to TrialNet. Britain has world class labs in immunology, same day sample movement, a single ethics/regulatory agency, access to patients previously identified from 150 hospitals, pre-existing research networks, and the National Health Service (NHS), which is the country’s sole healthcare provider and keeps good records. The group will support high-quality immunotherapy studies, it will be able to be both logical and strategic in setting the program, and it will be able to carry out replication of results, optimization of protocols and combination therapies (which is not what people are necessarily funding today). It will learn from failure and success, develop biomarkers, and increase the capacity for rapid progress.

Technology – Closed Loop

Roman Hovorka, PhD (University of Cambridge, Cambridge, UK)

The brilliant Dr. Roman Hovorka presented an overview of the closed loop in type 1 diabetes along with results from his recent work. As discussed at ATTD last month, Dr. Hovorka has data on thousands of patient days and nights (almost 12 years!) on closed loop and is currently conducting unsupervised outpatient trials. The results show great improvement at night, but less spectacular results during the day because of the meal challenges. Dr. Hovorka is satisfied that the technology works – he is now thinking about psychosocial aspects, commercialization, patient selection (which people will benefit?) and reimbursement.

• Commencing with an overview of insulin replacement therapy, Dr. Hovorka commented specifically that infusion sets and catheters are an area that needs work and that Abbott’s Libre is transforming glucose monitoring, even though it is not reimbursed.
• In Dr. Hovorka’s impressive unsupervised outpatient closed loop work, results show dramatically reduced variability overnight, usually reaching target by morning. Daytime variability is better than open loop, but not as dramatically so, as the system has to cope with frequent meals. The trials include unsupervised home use, without restrictions; patients bolus for meals and participants do whatever they want, including travel. The system has also been used in pregnancy and with younger children. Currently the team has 2,000 nights and 1,800 day-and-nights of experience across multiple randomized controlled trials. An ITT analysis showed lowering of average glucose, a 20% improvement in time in zone overnight, a 16% reduction in hyperglycemia and a 35% reduction in time spent in hypoglycemia. These results are quite impressive to us, as there are few therapies that can produce simultaneous improvement across all aspects of glucose control (like a pancreas transplant).
• Currently Dr. Hovorka uses the “Florence prototype”. This consists of the Abbott Navigator CGM transmitter, the Abbott Companion receiver (the size of smartphone), a control algorithm device (a tiny, book sized computer) and a Dana R insulin pump.
• There is significant inter- and intra-patient variability in required insulin dosing. Each person is different and each night is potentially different as well, which is. why conventional efforts often fail. Dr. Hovorka showed a chart of controller effort - the percentage of insulin dosed by the algorithm relative to open loop programmed basal levels. This varied tremendously for many patients, although there were a few that were relatively stable night after night, who would do well on open loop control.
• The patient experience was very positive. The key positive themes were reassurance, peace of mind, confidence, safety, improved control, sleep, not having to think about it, time off from diabetes demands, better control, and feeling better during the first half of the day. Negative themes were calibration difficulties, size of equipment, accuracy/trust, and frustration when equipment fails. The data was collected by Dr. Katharine Barnard (Southampton University, Southampton, UK).
• We are now seeing the incremental introduction of the technology into clinical practice. Medtronic has a threshold suspend product (MiniMed 530G) and has just launched a predictive low glucose suspend product (MiniMed 640G) beginning in Australia. Future steps might include treat to range, overnight closed loop, meal control with manual boluses (Medtronic’s 670G may be two to three years away from market, but it is great that an important player is showing their cards), meal and exercise control with automated dosing, and dual hormone closed loop as demonstrated by the Boston group led by Drs. Ed Damiano and Steve Russell. This approach can potentially derive better control, but it requires stable glucagon and the development of dual chamber pumps. However, there are those who argue that the addition of glucagon to closed-loop systems does not add significant benefits over insulin alone – Dr. Moshe Phillip (Schneider Children’s Medical Center, Petah Tikvah, Israel) made this point recently at ENDO. Overall, we feel that the two camps have mostly been talking past each other, and we hope to see more open debate (or ideally a head-to-head trial) on this issue in the future.
• Future challenges are no longer related to sensor accuracy/reliability, but to speed of insulin action, selection of users who will benefit, and the form of the technology (connectivity, instrumentation, integration). There is also the question of reimbursement – can we prove enough benefit to justify the cost? For this we are going to need larger trials with a focus on healthcare economics.

Transplantation

Paul Johnson, MD (University of Oxford, Oxford, UK)

Dr. Paul Johnson, a transplant surgeon at Oxford, reviewed the latest on whole pancreas transplants (particularly simultaneous pancreas and kidney [SPK] transplants) and islet transplants for diabetes. Whole pancreas transplantation has demonstrated good results, but it is still a major procedure that carries significant mortality risks, requires significant immunosuppression, is unlikely ever to be appropriate for children, and does not have a justifiable risk/benefit profile for hypoglycemia unawareness. Islet cell transplants have similar outcomes today, and are minimally invasive. But before they can become more mainstream, the field needs a renewable supply of islets and an approach that doesn’t require immunosuppression. Companies making significant progress in this area include Semma Therapeutics (a startup co-founded by Dr. Doug Melton [Harvard Stem Cell Institute, Cambridge, MA) and ViaCyte.

• Dr. Johnson made the case that transplantation is the ideal treatment for type 1 diabetes, because it has the potential to switch off ongoing cell destruction, replace destroyed islet cells, restore normoglycemia and reverse secondary complications. This might have been a ‘rose tinted spectacles’ perspective, but he made a good case, painting a favorable contrast with immunotherapy (which can’t replace islets) and technology (which can’t replace islets or switch off the autoimmunity).
• Whole pancreas transplants, particularly simultaneous pancreas and kidney transplants (SPK), can be very effective. At Oxford, the team has a 93% survival rate after one year (although we note that 7% is a significant mortality rate from a surgery). There is a 90% cumulative patient survival from 2002-2011. Only a third of patients have impaired glucose tolerance at this point, and 80-85% have very good glucose control after a year. There is also some evidence that there is a reversal of microvascular damage after time, although the jury is out on retinopathy.
• Islet transplantation is minimally invasive and currently offers similar glycemic outcomes to SPK. Everyone remembers the Edmonton protocol (2000) for islet transplantation, but there have been many steady improvements since then. The Edmonton group used to have 15% of patients insulin independent at five years – their current level is 62%. An important step was the use of campath as an induction agent. We have also seen data demonstrating more sustained insulin independence with the NIH Clinical Islet Transplantation Consortium’s CIT07 protocol compared to the Edmonton protocol. Dr. Johnson presented evidence that islet cell transplantation dramatically reduces hypoglycemia, A1c, variability, GFR and microangiopathy.
• However, with islets, there are still limitations. Islet transplantation is currently only used to treat small groups of highly selected patients, and to move to a broader population, the field needs to optimize islet isolation, improve initial graft survival, and find a renewable source of non-cadaveric islets. As a reminder, Dr. Doug Melton’s group at the Harvard Stem Cell Institute recently published a protocol for generating large quantities of insulin-producing cells from human stem cells; the group has formed a startup, Semma Therapeutics, that is working to translate the procedure into a clinical therapy. Another key challenge is the need to develop immunosuppressive-free strategies so that these therapies could be offered to children and generally pose less of a burden for patients.
• Dr. Johnson’s group has been testing MAILPAN (Defymed) an implanted bioartifical pancreas device that offers macroencapsulation of many islets to shield them from the immune system. ViaCyte is also an important player in macroencapsulation; the company recently provided an update on the ongoing phase 1/2 trial of its VC-01 cell replacement therapy at the UCSD Diabetes Public Forum. Other groups, including a team led by Dr. Daniel Anderson (MIT, Cambridge, MA) and supported by JDRF and the Helmsley Charitable Trust, are working on microencapsulation approaches that would encapsulate each islet individually.

New Therapies and Screening

Assessing the Effectiveness of DPP-4 Inhibitors on Glycaemic Control in Patients with Type 2 Diabetes: A Single Center Audit

Naim Desai, MD (School of Medicine, Liverpool University, Liverpool, UK)

This small observational study in a real-world setting showed that most patients taking DPP-4 inhibitors met goal and that broadly speaking, non-responders were discontinued after six months, in line with the NICE guidelines. This data builds confidence that DPP-4 inhibitors are effective in most cases, and that the NHS is not wasting money on non-responders.

• Type 2 diabetes currently accounts for 9% of the NHS budget, and people with diabetes have an average healthcare cost that is two to three times greater compared to the general population. Over the next 20 years, the costs of diabetes are projected to account for 17% of the NHS budget. The current spend is £2.1 billion/$3.1 billion on treatment and £7.7 billion/$11.5 billion on complications. Cost of drugs is around 15%.
• Against that background, DPP-4 inhibitors are considered expensive (between £26.60/$39.78 to £33.26/$49.74 for 28 days), so it’s important to gather real world data to assess their effectiveness. DPP-4 inhibitors are typically used when first line metformin and second line sulfonylureas fail. NICE guidance stipulates that DPP-4 therapy should only be continued if the patient experiences a >0.5% A1c reduction after 6 months. However, it’s not clear that patients always discontinue.
• A small observational study (n=36) in Liverpool showed that after six months of therapy, average A1c had dropped from 7.5% to 6.4%. Merck’s Januvia (sitagliptin) was the most commonly prescribed DPP4 inhibitor. 75% of the patients reaching target were appropriately continued, and 14% of the patients not reaching target were inappropriately continued. On closer review, some of those inappropriately continued met the target either at a later date or because A1c tests were late. This builds confidence that DPP-4 inhibitors are effective in most cases, and that the NHS is not wasting money on non-responders. 

TrialNet

Nicola Farthing, BSc (University of Bristol, School of Clinical Sciences, Bristol, UK)

TrialNet is an international consortium, and the UK arm is based at Bristol University. TrialNet UK has screened 6,000 at-risk people, of whom 4% are antibody positive. The consortium is conducting two prevention studies in the UK at present – oral insulin, anti-CD3, and abatecept.

• Bristol is the UK clinical center for Type 1 Diabetes TrialNet, an international consortium working to prevent or delay type 1 diabetes. There are 19 additional  screening sites across the UK. TrialNet refers patients at high risk of type 1 diabetes and screens adults and children for antibodies. Antibody positive patients are then invited to take part in trials. TrialNet currently is conducting two prevention interventions in the UK - oral insulin and abatecept.
• TrialNet has screened over 6,000 people in the UK to date, of whom about 4% have been found to be antibody positive. Most participants (56%) are adults over 20, and 22% are teenagers.
• The prevalence of islet antibodies is inversely related to age. Results suggest that to find 100 participants for trials, TrialNet would need to screen 4,000 children less than 12 years or 5,100 teenagers or 10,500 adults. To build a highly effective recruitment would require more scale and speed. The low (4%) yield is a major cost factor. This is also an issue facing the T1DUK Immunotherapy Consortium that was announced yesterday. It wasn’t clear whether TrialNet would work in partnership with this new initiative.

Two Year Observational Studies of GLP-1 Agonists

Margaret Jackson (Dudley Group of Hospitals NHS Foundation Trust, Dudley, UK)

Although NICE only recommends continuing GLP-1 agonists after six months if A1c and weight targets are met, this study aimed to investigate whether the value was sustained beyond six months in a real-world setting. In this small observational study of n=141 patients (who didn’t discontinue therapy), average A1c and weight reduction at six months were perfectly sustained out to two years. Scatter plots showed a wide variation in response to therapy, but 27% of patients were able to achieve and sustain the NICE guidelines (>1% A1c reduction and >3% body weight reduction). Despite the guidelines, the other patients typically continued on drug if they were seeing some reasonable level of therapeutic benefit.

• According to guidelines, GLP-1 agonists should provide a >1% A1c reduction and >3% weight loss at 6 months. NICE guidance recommends discontinuing GLP-1 agonists if these targets are not met.
• In this observational study, n=141 patients taking GLP-1 agonists were followed out to two years. The average age at baseline was 56, duration of diabetes was about 10 years, BMI was 39 kg/m² and baseline A1c was 9.1%. 60% of patients took AstraZeneca’s Byetta (twice daily exenatide) and 40% took Novo Nordisk’s Victoza (liraglutide) – somewhat surprising given that Victoza held ~77% market share within the GLP-1 agonist class as of the end of 2014.
• At both six months and two years, average A1c reduction from baseline was 0.6% and weight reduction was 4 kg (3.5%). 37% of patients had achieved an A1c drop of 1%, 51% had a weight loss greater than 3%, and 27% achieved both goals. A scatter plot showed a broad variation in response to therapy across the dimensions of A1c and weight.
• A multivariate analysis showed that the key predictors of success at two years were early A1c reduction, higher starting A1c and a shorter duration of diabetes. Around 80% of patients with an A1c reduction of >1% at six months, baseline A1c >9.5% and duration of diabetes <5 years maintained their A1c reduction at two years. As a result, Dudley is now selecting patients more carefully. However, the group typically does not discontinue GLP-1 agonist therapy if some level of therapeutic benefit is observed, even if the patient doesn’t meet the NICE targets, for fear that the patient would be even worse off without the drug.

Non-Insulin Adjunct Therapy in Type 1 Diabetes

Metformin

John Petrie, MD, PhD (University of Glasgow, Scotland, UK)

In this first of three lectures on non-insulin adjunct therapy in type 1 diabetes, Dr. John Petrie considered the case for using metformin. He noted that insulin is limited by hypoglycemia and drives weight gain, so there is room for an adjunct that can lower or maintain A1c with the same or lower rates of hypoglycemia while mitigating weight gain and insulin dose (and ideally improving cardiovascular outcomes!). Since metformin has so many benefits in type 2 diabetes and is safe and inexpensive, it’s surprising that current evidence with type 1 diabetes is actually quite thin. Fortunately, there are a number of randomized controlled trials underway. The REMOVAL trial (n=500), completing in 2017, will measure micro- and macrovascular outcomes and markers.

• Scotland has poor A1c performance compared to many countries – only 20% at goal (<7.5%). Insulin is the drug of choice, but has the downsides of hypoglycemia and weight gain. In the DCCT, those in the intensive group gained significantly more weight. In Scotland, life expectancy for adults aged 20 with type 1 diabetes is 11 years less for men and 13 years less for women compared to those without diabetes, and the key mortality factor for older people is cardiovascular disease. The ideal adjunct therapy would therefore lower or maintain A1c, eliminate hypoglycemia, lower the insulin dose (and hence weight gain), and reduce cardiovascular complications.
• Metformin is widely recommended as first line therapy for type 2 diabetes and has a number of beneficial actions that could possibly apply in type 1 diabetes as well. In type 2 diabetes, metformin decreases hepatic glucose production, increases insulin sensitivity, decreases LDL cholesterol, reduces weight, has direct vascular effects (on the endothelium), is anti-inflammatory, anti-thrombotic, and leads to decreased markers of oxidative stress. In UKPDS, metformin also showed potential for cardiovascular benefit.
• Small trials of metformin in type 1 diabetes show evidence that adding metformin can lead to lower insulin doses. A meta-analysis of small type 1 diabetes trials didn’t show a statistically significant A1c reduction but did show evidence of weight reduction and lowering of insulin dose. Strangely, in another n=99 trial, severe hypoglycemia increased, which was explained by not reducing insulin dose enough in the metformin arm. We saw similarly mixed results at ENDO from a T1D Exchange study (n=140) investigating metformin as an adjunct to insulin in overweight adolescents with type 1 diabetes; the study did not meet its primary endpoint of demonstrating greater A1c reductions vs. placebo but did show promising results with regard to insulin dose and BMI z-score.
• The REMOVAL trial is a three-year international multi-center randomized controlled trial (funded by JDRF) investigating metformin as an adjunct to insulin in people with type 1 diabetes over 40, with a focus on cardiovascular risk. Participants have three or more cardiovascular risk factors, and the primary outcome is carotid intima-media thickness (CMT). Subjects are also assessed for retinopathy. It’s a treat to target trial, aiming for an A1c of 7.0% in both the placebo and metformin arms. It finished recruiting nearly 500 participants in 2014 and will conclude in 2017. There are also other ongoing clinical trials of metformin in type 1 diabetes, notably T1DExchange, (young adults) which should report at ADA 2015, EMERALD (finishing 2018), and Adelaide. Most will assess macrovascular health, although there are no cardiovascular outcomes trials.

GLP-1

Sten Madsbad, MD, PhD (Hvidovre Hospital and University of Copenhagen, Copenhagen, Denmark)

There is strong evidence that GLP-1 agonists work very well alongside insulin in type 2 diabetes. In type 1 diabetes, we see broadly the same results – better control, lower insulin dose and weight loss, at the expense of some GI side effects. However, results are particularly good for type 1 diabetes patients with some residual beta-cell function, with some patients actually discontinuing insulin (some people in the audience were concerned about the safety of this). The hope is that GLP-1 might be protective in type 1. A study (ClinicalTrials.gov Identifier: NCT0189917) investigating Novo Nordisk’s Victoza (liraglutide) in newly diagnosed patients is currently recruiting participants, and the Novo Nordisk-sponsored ADJUNCT ONE  (treat to target; NCT01836523) and ADJUNCT TWO (fixed insulin dose; NCT02098395) trials are expected to complete this year.

• GLP-1 agonists could potentially fill several important unmet needs in type 1 diabetes. Conventional insulin therapy does not match peak glucose excursions, there are concerns about hyperglycemia and hypoglycemia, glucagon is not addressed, and more type 1 diabetes patients are becoming obese, partially driven by insulin.
• Dr. Madsbad discussed several studies in which adding Novo Nordisk’s Victoza (liraglutide) to insulin in type 1 diabetes allowed for a lower insulin dose and weight loss while maintaining or improving glycemic control. Treatment effects were greater for patients that were C-peptide positive (were still making some insulin). Time in zone was better than with insulin alone  - and some subjects even discontinued insulin altogether. Typical GI side effects of GLP-1 were observed in these studies. Interestingly, Dr. Madsbad commented that weight loss was double what might have been seen with type 2 patients.
• Ongoing studies investigating liraglutide in type 1 diabetes include NewLira, ADJUNCT ONE, and ADJUNCT TWO. NewLira (sponsored by Hvidovre University Hospital; currently recruiting) will investigate the effect of liraglutide vs. placebo in 80 newly diagnosed patients; the primary endpoint is change in beta cell function at 52 weeks. ADJUNCT ONE (sponsored by Novo Nordisk) is a treat to target add-on of liraglutide or placebo to insulin for 52 weeks. It enrolled n=1398 patients and will finish in May 2015; the primary endpoints are change in A1c, body weight, and total daily insulin dose. ADJUNCT TWO (also sponsored by Novo Nordisk) is a 26-week fixed dose insulin trial with three doses of liraglutide that enrolled n=835 patients and will conclude in April 2015; the primary endpoint is change in A1c.

Options Other than Metformin and GLP-1

Parth Narendran, MD, PhD, (University of Birmingham, Birmingham, UK)

Dr. Parth Narendran covered a wide range of agents, including exercise, that could be used as adjuncts to insulin in type 1 diabetes, though the quality of the data is generally poor at this stage. He concluded that the most interesting combinations are SGLT-2 inhibitors and GLP-1 agonists. One of the most interesting slides he showed was CGM of a patient before and after adding an SGLT-2 inhibitor. It appeared that there was a marked improvement in time in zone, particularly in hyperglycemia, which went from 12% to 2%. This is consistent with the mechanism of action of SGLT-2 inhibitors. Both AstraZeneca’s Farxiga (dapagliflozin) and J&J’s Invokana (canagliflozin) are currently being investigated in type 1 diabetes. We have long believed that this class could hold great appeal for that population due to the potential for reduced insulin doses and weight loss as well as reduced glycemic variability; however, the recent anecdotal reports of patients experiencing euglycemic DKA certainly warrant further investigation – this is going to be a big issue on the safety front due to risk, unless those on SFUs start testing ketones (regularly? We aren’t sure yet) with the Abbott kit (a bit ironic in light of their BGM moving to FreeStyle Libre, where it is approved).

• Adjunct studies of DPP-4 inhibitors in type 1 diabetes suppressed glucagon and didn’t cause hypoglycemia, but did not show any compelling impact on A1c or weight.
• Adjunct studies of AstraZeneca’s Symlin (pramlintide) showed a reduction in A1c, a weight improvement, and a reduction in insulin dose. However, many patients have issues with taking three additional injections per day, and Symlin is difficult to dose. Nausea is self limiting, and intolerable in 20%, and there has been increased hypoglycemia when the insulin was not reduced aggressively enough. The drug is also not licensed in the UK.
• Alpha glucosidase inhibitors slow down the absorption of carbohydrate. Adjunct studies show that there is a postprandial decrease in glucose, but up to 75% of patients have GI side effects.
• There are ongoing proof of concept studies for SGLT-2 inhibitors in type 1 diabetes. In trials, AZ’s Farxiga (dapagliflozin) and Lilly/BI’s Jardiance (empagliflozin) showed a modest or non-significant reduction in A1c, and anecdotal CGM data shows much less glycemic variability, but it is early days yet.
• Exercise is likely to have a benefit, but a meta-analysis of n=452 patients didn’t offer compelling proof. It is known that high intensity training improves insulin sensitivity.
• In summary, SGLT-2 inhibitors and GLP-1 agonists appear very promising, even though they are not yet indicated for use in type 1 diabetes. Exercise and pramlintide are already approved for that indication, and metformin, pioglitazone, acarbose, and DPP-4 inhibitors are not approved but potentially of use in selected patients.

The Future of Diabetes in the NHS

The Future of Diabetes in the NHS

Simon Stevens (CEO, NHS England, Leeds, UK)

Simon Stevens, the CEO of NHS England, gave a very impressive keynote presentation, introduced by the admirable Baroness Young. NHS England is the sole health provider in England, with an annual budget of £113 billion ($169 billion) and over 709,000 employees. Mr. Stevens gave a pithy and lucid presentation focused on the launch of the new National Diabetes Prevention Programme for England. This program is designed to enroll thousands of at-risk obese men and women and provide DPP-like services to reduce their progression to type 2 diabetes. A task force hopes to enroll 10,000 people by the end of the year, while building capabilities for a much wider effort. About a quarter of British adults are obese (that’s about 12.5 million people). Public Health England (PHE) asserts that an estimated 9.6 million Brits are at high risk of developing type 2 diabetes.

• “Obesity is the new smoking”. Currently, 26% of men and 24% of British women are obese. Obesity rates have doubled over the last two decades. One in five children are obese by sixth grade.
• “We are sleepwalking into a tidal wave of new illness”. The principal risk factor for type 2 diabetes is obesity. In the UK, type 2 diabetes costs £9 billion ($13.5 billion)and leads to 20,000 early deaths per year. The NHS is expecting a population of over four million people with diabetes within 10 years. It has not been able to reduce or limit prevalence with its current policies, although there has been some impact on complications.
• Mr. Stevens announced the launch of first NHS National Diabetes Prevention Programme. The landmark Diabetes Prevention Program (DPP) trial in the US showed that it is possible to reduce the risk of progression from prediabetes to diabetes by 58%. The UK program will have many of the same elements, although Mr. Stevens implied that it would employ more recent learnings in behavioral aspects of diabetes and would rely more on groups (the DPP had a lot of individualized coaching). He noted that the UK is not “reinventing the wheel” on intervention, and that the key issues are implementation and scaling.
• At launch, the Programme has three main aspects: to get off to a quick start, to develop the optimal program delivery and roll out to more providers quickly, and to launch a prevention program within the NHS itself. Over the next 12 months, Mr. Stevens is targeting enrollment of at least 10,000 high risk individuals at initial “demonstrator” sites. Subsequently, more providers will be established and trained until the Programme has broad coverage. A procurement notice is going out this week. Finally, the NHS is launching an internal program for NHS staff themselves to demonstrate that they take the idea of a healthy workforce seriously. The obliging British press reported this initiative as “Fat doctors told to attend Slimming World.”
• Seven “demonstrator” sites will seek to enroll 10,000 high-risk people in the initial phase. The initial phase will provide programs for weight loss, physical activity, cooking and nutrition, and peer support, plus telephone and online support from trained professionals. The demonstrator sites will monitor and test their programs to help design and roll out the national program.
• As a final note, Mr. Stevens put Prevention into the context of a five-year plan for the NHS, which also contained an initiative to redesign services and give more control to the people who benefit from them. The NHS has various ‘silos’ (e.g. GP services, mental health services), which will be redesigned to avoid fragmentation and provide a more seamless and coordinated delivery to the customer. It’s also expected that this will involve more use of technology to communicate with patients. Finally, patients will be getting more direct control of their services – a shift towards recognizing that NHS will only be successful if clients have more say. To take one recent example, feedback from children has led to pediatric outpatient services being reorganized around school locations and hours.

Questions and Answers

Q: Will the politicians meddle and mess this up?
A: I will say two moderately diplomatic things. First, it’s only appropriate in a democracy that elected representatives are interested in policy for something that matters a lot to the electorate. Secondly, for some time now we have been setting out the agenda for the NHS and making the case for the five -year plan, of which this announcement forms a part. Hopefully this will be durable.

Poster Session

If We Followed NICE Guidelines and Treated People With Diabetes Earlier, Patients and the NHS Would Benefit From a Reduction in Complications, Mortality and Direct Healthcare Costs.

Mike Baxter et al (Sanofi, Guildford, UK)

A simple poster, sponsored by Sanofi and JDRF, modeled the healthcare economics of lowering A1c according to NICE guidelines for escalating therapy for type 1 and type 2 diabetes. Using the IMS Core Diabetes Model, the authors calculated that there was a potential cost savings (in avoided complications over a 25-year period) of £995 million/$1.5 billion in type 1 diabetes and £4.5 billion/$6.7 billion in type 2 diabetes. They didn’t model the incremental cost of therapy, but asserted that it would be lower than the cost of complications. As the focus often turns to the high cost of diabetes drugs, the manufacturers make the valid point that it is still cheaper and better for society to treat early and maintain a lower A1c.

Diabetes UK Research

Autoimmunity: Measuring and Modifying the Immune System in Type 1 Diabetes.

Timothy Tree, PhD (King’s College London, London, UK)

Dr. Timothy Tree did a terrific job of explaining the role of the immune system in type 1 diabetes in only 30 minutes. There are a number of distinct changes in the immune system in patients with type 1 diabetes compared to healthy people. These include autoantibodies, increased cytotoxic T-cells, increased helper T-cells and B cells, and less effective regulatory T-cells. On top of this, there is enormous inter-patient heterogeneity, meaning that the goal of immunotherapy should be to understand why some therapies work for some people and not others and to tune therapies to fix only the part that is wrong for each individual patient.

• The immune system has to balance immunity and tolerance. It has great diversity and specificity expressed throughout numerous innate and adaptive cells. People can make over a billion T-cell receptor shapes, and in everyone’s T-cell repertoire there will be cells with receptors that are autoreactive. Regulatory T-cells (Tregs) prevent the destruction of self-tissue by autoreactive T-cells and provide a balance between destroying pathogens and hurting oneself.
• Islet cell autoantibodies identify people with type 1 diabetes. The key ones are (pro)-insulin, GAD 65, IA-2, and ZnT8. They don’t directly cause diabetes, but they are very good markers of future disease. People who have three or four autoantibodies will almost certainly develop diabetes in the future.
• Type 1 diabetes is characterized by a range of detectable immunological changes. T-cells that produce IFN-γ (a proinflammatory cytokine) are the main cytotoxic T-cells that appear in response to islet antigens, and lead directly to beta cell death. In contrast, IL-10+ regulatory T-cells are decreased in type 1 diabetes. These cells will indeed stop islet apoptosis if enough of them are present, opening an interesting avenue for new therapies. There is also a decreased CD25+ regulatory T-cell response in type 1 diabetes. The Tregs work much less effectively than expected, which is associated with poor sensitivity to IL-2 (a growth factor). 
• Type 1 diabetes is highly heterogeneous across patients, making it hard to develop one-size-fits-all therapies. Under the umbrella of the new T1DUK Immunotherapy Consortium, the team wants to understand why some therapies work for some people and not others. Ultimately the goal is to tune immune therapies to fix only the part that is wrong for each individual patient.

The CTLA-4 Pathway in Type 1 Diabetes

Lucy Walker, PhD (University College London, London, UK)

Dr. Lucy Walker, with support from Diabetes UK, has been investigating the role of the CTLA-4 pathway, which has been strongly linked to susceptibility to autoimmune disease. In a mouse model, CTLA-4 turns out to have a regulatory effect on the immune system and suppresses the production of follicular helper T-cells that are overrepresented at the site of autoimmune attacks. Her mouse model also correlates well with what happens in humans with diabetes. Dr. Walker’s long-term goal is to develop drugs that mimic the effects of regulatory T-cells that might swing the balance of the immune system back far enough to reverse autoimmune effects.

• The T-cell regulatory gene CTLA-4 has been linked with a susceptibility to autoimmune disease.  T-cells expressing CD28 and CTLA-4 bind similarly to antigen-presenting cells (APCs), but CD28 promotes T-cell activation and can drive autoimmunity, while CTLA-4 has a regulatory function (working in the opposite direction). Furthermore, CTL4-A is expressed at high levels in regulatory T-cells (Tregs).
• Injecting Tregs into a type 1 diabetes mouse model successfully prevented the occurrence of diabetes. Furthermore, knocking out CTLA-4 from the Tregs canceled out their effect. The Treg cells transfer binding sites from the target cell and actually absorb them, stripping them of ligands that could mediate autoimmune cell destruction.
• It turns out that Tregs are using CTLA-4 to limit CD28 signaling. The balance between CD28 and CTL4-A controls the propensity of T-cells to differentiate into follicular helper T-cells (Tfh). Tfh are found in mouse models at high levels at the site of the autoimmune attack, and they are also overrepresented in people with diabetes, which serves to validate the mouse model.

Rank Nutrition Lecture

Modulating Gut Hormones To Treat Obesity And Type 2 Diabetes: Fact Or Fantasy?

Rachel Batterham, MD, PhD (University College London, London, UK)

Leading obesity researcher Dr. Rachel Batterham asserted that gut hormones like PYY, ghrelin, and GLP-1 are excellent targets for therapies for type 2 diabetes. She explained that gut hormones don’t return to normal levels after weight loss, making maintenance more difficult. It further seems that the excellent results of bariatric surgery rest more upon hormonal effects than restriction or patients’ efforts. So the key question becomes ‘how do we get the effect of surgery, without the surgeon?’

• The gut is the largest endocrine organ and its primary role is to let the brain know what has been ingested. It does this via a selection of hormones that regulate energy and glucose homeostasis, including CCK, GIP, ghrelin, Apo AIV, GLP-1, GLP-2, oxontymodulin, PYY, FGF-19, and neurotensin – not forgetting insulin, glucagon and amylin. In this talk, Dr. Batterham focused on PYY, ghrelin and GLP-1.
• PYY is a hormone produced by L cells in the gastrointestinal tract. Humans infused with PYY decrease their food intake by 25% compared to control groups. PYY also works in obese people, for up to 24 hours. But obesity in humans (and mice) is associated with low circulating levels of PYY – so there appears to be a bi-directional relationship between obesity and PYY. PYY acts mostly on the reward center of the brain and it also improves glucose tolerance in the islet cells. It is also present in saliva and modulates taste.
• Ghrelin levels go up before eating and drop during feeding. Administration of ghrelin leads to increased hunger and food intake. Ghrelin also acts on the reward centers of the brain and influences taste. Inhibiting ghrelin is therefore a great pharmaceutical opportunity, and there is currently great activity in developing ghrelin O-acyltransferase (GOAT) inhibitor enzymes to achieve this.
• GLP-1 increases insulin production, decreases glucagon, improves satiety and reduces appetite, and leads to weight reduction. As a reminder, Novo Nordisk’s Saxenda (liraglutide 3.0 mg) is the first GLP-1 agonist approved for the treatment of obesity in the US and EU. GLP-1 modulates the appetite and reward areas in the brain; Dr. Batterham shared data from fMRI experiments showing that GLP-1 blocked the effects of food images in the reward areas of the brain. This GLP-1 response was reduced by 25% in women with type 2 diabetes or prediabetes, but there was no difference in men. GLP-1 response was reduced by 8% in overweight people and by 20% in obese people, compared to normal weight controls. People have differences in GLP-1 response; 67% of the response is heritable, and it can be modulated by obesity or fat in the liver.
• Maintaining weight loss after dieting is a real issue partially because of the effects of gut hormones. Studies showed that a year after dieting, hunger scores didn’t go back to baseline. Dieting and weight loss cause a marked increase in circulating ghrelin levels and a decrease in PYY.
• It seems likely that high protein diets are best for the achievement of sustainable weight loss, particularly in people at genetic risk for obesity.  Diet studies are usually not long enough and suffer from dropouts, so Dr. Batterham tested isocaloric diets with varying macronutrient proportions. She was able to show that high protein diets yield lower hunger scores, more PYY and less ghrelin than alternatives. It was noted that in subjects with gene variants linked to obesity risk (FTO gene), hunger and ghrelin levels don’t suppress as much after a meal. People with at-risk variants also had very different brain responses to food images (including greater activation in incentive areas), even if they weren’t particularly obese (yet). Dr. Batterham concluded that a high protein diet was the best option for those with FTO variants that put them at high risk for obesity. For the general population, there is currently no strong evidence favoring particular macronutrients over others, largely due to the poor quality of most nutrition studies (e.g., observational design, self-reported data).
• There is a remarkable overlap between the effects of gut hormones and bariatric surgery, likely because the effects of surgery have more to do with biology than with restriction or compliance. So the key question is “how to reproduce the surgery without the surgeon?” Bariatric surgery mainly reduces caloric intake through reduced appetite, changes in food preferences, and a reduced interest in reward value of food. Notably, GLP-1 and PYY levels increase markedly after surgery. Like the response to GLP-1, the weight loss after surgery is normally distributed and we can’t tell in advance who will do well or not. But we do know that poor responders have lower PYY and higher ghrelin after surgery. Dr. Batterham believes that success with bariatric surgery is therefore a result of biology, not of patients’ efforts. Dr. Lee Kaplan (Massachusetts General Hospital, Boston, MA) has reached similar conclusions based on his extensive research into the mechanisms of bariatric surgery; he has also discussed the impact of changes in gut microbiome composition following surgery.

Plenary: Adipose Tissue – You Can Have Too Much Of A Good Thing

Bariatric Surgery As A Treatment Option In Diabetes

Nick Finer, MD, PhD (University College London, London, UK)

Bariatric surgery expert Dr. Nick Finer made a persuasive case that many more patients with diabetes should be considered for surgery, since it is more effective than conventional medical therapy. He cited the impressive diabetes remission rates, reductions in mortality, and cost savings with the procedure as evidence that greater use of bariatric surgery would be beneficial for patients and for the healthcare system as a whole. We have heard a number of speakers in the US similarly argue for expanded use of bariatric surgery in patients with diabetes; Dr. Philip Schauer (Cleveland Clinic, Cleveland, OH) has been a particularly outspoken proponent. However, the argument is clearly far from settled, as evidenced by the debates at ACC this month over how to balance the risks of malabsorption and other long-term complications with the metabolic benefits of surgery.

• Only 21% of patients currently meet all the targets for diabetes and metabolic syndrome, implying that bariatric surgery should have a bigger place in clinical treatment. Diet and exercise deliver therapeutic benefits five years after intervention in only 5% of cases.
• Bariatric surgery has three main variants in the UK – in order of increasing perioperative risk, these are gastric band, sleeve gastrectomy, and gastric bypass. The number of bariatric surgeries worldwide is increasing dramatically - except in the UK. Sleeve gastrectomy and gastric bypass are also increasing at the expense of gastric band. NICE guidelines encourage the use of bariatric surgery for new onset type 2 diabetes in obese patients, but they are clearly not being followed.
• Complete remission rates with bariatric surgery are somewhat lower than generally recognized. They are about 5% with the gastric band and 25% with Roux-en-Y gastric bypass surgery (RYGB). Complete remission is defined as normoglycemia for a year or more, without medication. Overall remission rates (defined as no longer meeting the glycemic criteria for diabetes) at two to five years are around 50-70%. But remission is not exactly the correct term, since bariatric surgery does not restore normal physiology and gives rise to supra-physiological levels of hormones, leading to significant glucose excursions. A study shows that post-RYGB patients spend 12% of the time in hypoglycemia and 5% in hyperglycemia each day.
• Bariatric surgery is literally life saving for some. There is a halving of total mortality after ten years (after an initial excess of peri-operative mortality). Randomized controlled trials like STAMPEDE have shown that bariatric surgery wins hands down on A1c compared to best practice medical therapy. Meta-analyses show a 1.5% fall in A1c on average. On the other hand, Dr. Timothy Garvey (University of Alabama, Birmingham, AL) argued at Obesity Week that “weight loss works no matter how it is achieved” and cautioned providers against automatically jumping to the conclusion that bariatric surgery is the most effective way to achieve type 2 diabetes remission.
• The drawbacks of bariatric surgery include uncertainty over ultimate weight loss, resolution of microvascular complications in only a minority of patients, post-gastric bypass hypoglycemia (late dumping) in about 0.2% of patients, and long-term vitamin deficiencies. There is a broad, normal distribution of weight loss after surgery, which doesn’t appear to be related to patient efforts. About 10-20% of patients experience improvements in preexisting retinopathy, a similar proportion progresses, and the rest remain stable.
• Repeated economic analyses demonstrate that bariatric surgeries are economically “dominant” (produce savings greater than the cost of the procedure) in only three to four years. However, only 0.6% of the eligible population in the UK population has been treated to date, so the NHS would need significant organization to take advantage of these benefits on a larger scale.

Brown Adipose Tissue Is A Therapeutic Target In Humans

Sven Enerback, MD, PhD (University of Gothenburg, Gothenburg, Sweden)

Brown adipose tissue (BAT) is regarded as the ‘good’ fat. It is visible in rodents and human infants between the shoulder blades, and another form is interspersed with the normal white adipose cells. BAT is activated by cold temperature, and one of its jobs is to create heat, but it is also an endocrine organ. It can decrease insulin resistance and cause weight loss, making it an interesting drug target. BAT has been a fairly hot topic in obesity research that has produced several intriguing findings in recent years. As one example, a Harvard Stem Cell Institute team recently identified two compounds that can convert white fat to brown fat in human cells; the researchers are currently in discussions with several pharmaceutical companies about continuing the work.

• There are two types of brown adipocytes. Classic (normal) brown adipocytes (fat cells found between the shoulder blades in rodents) and white adipocytes come from unrelated precursors. “Recruitable” brown adipocytes (beige/brite adipocytes) are dispersed among white adipocytes and come from a similar source.
• Brown adipocytes help to protect against diabetes and fractures in rodent models. Brown adipose tissue (BAT) is protective of diet-induced type 2 diabetes in mice. Mice with a lot of brown fat (induced via the FOXC2 gene) have better insulin sensitivity and higher bone mass. Browner bone marrow adipocytes seemed to create more bone, implying they act as endocrine organs.
• There is also MRI evidence for brown fat in humans that behaves similarly. Human BAT is also activated by cold outdoor temperature, leading to decreased fat mass and improved insulin resistance. Babies have interscapular brown fat, which they lose with age - but the remaining beige/brite fat still exists and can be activated by cold, insulin, and certain drugs. Dr. Enerback presented data showing that after giving small amounts of insulin, glucose uptake was very high in activated brown fat.

-- by Emily Regier and John and Kelly Close