Hello from a grey and windy Liverpool, as we look out across the River Mersey following the 2014 Diabetes UK Professional Conference. As ever, the conference squarely focused on the patient, with ample practical material for the clinician. Also keeping with last year, we celebrated some recent drug therapy launches – in this case it was J&J’s SGLT-2 inhibitor Invokana (canagliflozin), which was launched in the UK just before the conference began (Feb 24, 2014), and Takeda’s DPP-4 inhibitor Vipidia (alogliptin; branded as Nesina in the US), which was launched at the conference. Such is the public consciousness of diabetes in the UK, that a leading national newspaper, the Daily Express, ran a front-page headline about Invokana: “Daily Pill to Beat Diabetes” on Feb 25th. Also of interest was a short speech from Baroness Young, Chief Executive of Diabetes UK, outlining the many patient-centered initiatives that Diabetes UK have successfully spearheaded over the last 12 months. These include a major national advertising campaign for diabetes screening (to ‘find the 7m people in the UK at high risk’), a hotline (Careline) staffed with professionals for patients needing advice, and a partnership with a leading retailer, Tesco, to raise money (over £10 million, or $17 million) and awareness. There has also been success on the legislative front - UK schools will now be required by law to provide appropriate support for pupils with diabetes. Diabetes UK is helping draft guidance for these support programs.
The conference itself didn’t have much in the way of new data, but there was clear interest in SGLT-2 inhibitors and a lot of focus on drug safety. It was surprisingly quiet on the device front, but we did see some interesting pump audit data from Britain’s National Health Service (NHS), as well as full sessions from Abbott (on Flash Glucose Monitoring and ambulatory glucose profiles) and Animas (on the new pediatric indication for the Vibe).
Our full conference coverage is included below, organized into four sections:
1. Diabetes Drugs
Ensuring Safety with New Agents
Clifford Bailey, PhD (Aston University, Birmingham, UK)
Dr. Bailey appeared skeptical of the current regulatory environment for ensuring cardiovascular (CV) safety. He definitely believes that there are lingering CV safety issues with most anti-diabetic drugs, yet in most cases the absolute risks of the treatments are much smaller than the risks of the disease. Nonetheless, interpretation of trial results based on limited numbers, populations, and timescales is fairly difficult, and we need to be accepting of revisions without prejudice as new information emerges. The legacy of the FDA’s CV safety guidance is that many very large-scale trials are being conducted both pre- and post-approval, typically in high-risk populations. Dr. Bailey also felt that the EMA was ‘less obsessed’ with the numbers than the FDA when it came to assessing whether a drug was likely to be safe.
- According to Dr. Bailey, many anti-diabetic agents (even the veterans of the field) have lingering cardiovascular (CV) safety concerns that we don’t fully understand. However, it must be remembered that on absolute risk basis, the risks of adverse effects from drug therapy are generally much smaller than the risks of diabetes complications such as MI and stroke.
- Regulatory decisions based on limited numbers, populations, and short-duration studies are difficult, as we saw from the Avandia (rosiglitazone, GSK) issue. The notorious meta-analysis from Dr. Steve Nissen (Cleveland Clinic, Cleveland, OH) suggested there was a CV safety signal (even though there were concerns with the methodology – such as trials with no events being excluded). Interestingly, analysis of the individual studies revealed that longer trials showed reduced mortality compared to shorter ones. The larger study, RECORD, showed no significant difference in risk over multiple years, and after two analyses of the data, the FDA has lifted its prior restrictions on the drug (read our report on the FDA’s 2013 Advisory Committee meeting on Avandia). According to Dr. Bailey, we need to accept revisions without prejudice as new information emerges.
- The legacy of this saga is that in December 2008, the FDA introduced guidance on CV safety, requiring long-term post-marketing study if the upper 95% confidence interval exceeded a 1.3 hazard ratio, and requiring additional pre-approval trials if the interval exceeded 1.8. Dr. Bailey cited the example of Novo Nordisk’s Tresiba (insulin degludec). In this case, the FDA re-defined the major adverse cardiovascular events (MACE) endpoint, and because of a lack of events in the submitted studies, the upper bound exceeded 1.8, requiring further safety work before approval. Dr. Bailey commented that “the EMA was not so obsessed on the numbers, and looked at the information a little more.” For the typical phase 2/3 program, many hundreds of events would be required to meet FDA guidance, even with hazard ratios close to 1.0 – implying the need for much larger, longer trials on higher risk patients.
- The FDA has now asked many pharma companies to conduct very large scale post-marketing cardiovascular outcomes trials, which should provide nearer to real-world information. Companies have also felt competitive pressure to conduct the trials. Typically there are important differences in enrolled populations, enrollment size, and time intervals, which make it difficult to draw direct comparisons.
Simon Heller, MD (The University of Sheffield, Sheffield, UK)
Dr. Heller covered two important large cardiovascular (CV) outcomes trials for DPP-4 inhibitors – SAVOR-TIMI 53, studying AZ’s Onglyza (saxagliptin); and EXAMINE, studying Takeda’s Nesina/Vipidia (alogliptin). But studies enrolled different at-risk populations. The trials both clearly showed no increase in CV risk, a (modest) improvement in glucose control, and no increase in the risk of pancreatitis and pancreatic cancer. The data on pancreatitis and pancreatic cancer is reassuring, but does not yet establish safety. In SAVOR, there was a significant and unexpected 27% increase in hospitalization due to heart failure – as a reminder, the FDA formally requested SAVOR data from AZ in February to evaluate the heart failure signal (read our report).
- Existing post-marketing surveillance for DPP-4 inhibitors has been reassuring so far, but it is still insufficient to establish safety. Underlying sources of concern with DPP-4 inhibitors include the fact that (by design) they increase GLP-1 receptor activation well above normal physiological levels, and that they have been tentatively linked with pancreatitis and pancreatic cancer in animal studies. In the phase 2/3 programs for alogliptin and saxagliptin there was no increase in cardiovascular risk, but the background CV risk was low, suggesting a need for longer-term studies.
- The SAVOR-TIMI 53 trial established that there was no increased CV risk for AZ’s Onglyza (saxagliptin) compared to placebo over two years of follow up in ~16,500 patients with multiple risk factors for cardiovascular disease. There was also no indication of any increased risk of pancreatitis or pancreatic cancer. In SAVOR TIMI, the study participants were ~65 years old on average, 33% female, and had a mean baseline A1c of 8.0%. Approximately 80% had established cardiovascular disease and the remaining 20% had multiple risk factors. After a median duration of 2.1 years, the hazard ratio for the CV composite endpoint was 1.0, with a composite event rate of 7%. A1c was 0.3% lower after two years in the saxagliptin group, and 40% of participants in this group achieved an A1c <7%, versus only 30% in the control group. The saxagliptin group also required less intensification of therapy.
- A surprise from SAVOR was that the relative risk of hospitalization for heart failure was 1.27 for saxagliptin, (which was statistically significant). However, the risk of CV death was not affected.
- The EXAMINE trial established that there was also no increased CV risk for Takeda’s Nesina/Vipida (alogliptin) in a study of 5,380 patients at very high risk of a CV event. There was also no increase in pancreatitis or pancreatic cancer, although incidence rates were low. These patients had “acute coronary syndrome” (ACS), meaning they were randomized on average just 44 days after a prior event (77% had had an MI, 23% had unstable angina requiring hospitalization). In EXAMINE, participants at baseline were 68% male, 73% white, mean duration of diabetes was seven years, mean BMI was 29 kg/m2, and mean baseline A1c was 8.0%. The Kaplan-Meier plot showed no statistical differences over time with a hazard ratio of 0.96 for the composite endpoint. Alogliptin also showed a modest improvement in A1c.
- A composite of hospitalization due to heart failure and CV death showed no excess risk with alogliptin. In EXAMINE, CV mortality and hospitalization for heart failure should be analyzed together, since some of the deaths are due to heart failure. However, the trend is to reduced mortality and increased risk of hospitalization for heart failure.
- The trials focus on slightly different populations, but both show no increased CV risk. The lack of data on pancreatitis and pancreatic cancer is reassuring but doesn’t establish safety. One difference in the two trials was increased hospitalization for heart failure with saxagliptin, but this could be due to reduced power in EXAMINE or the different study population. Other trials might shed light on this.
Miles Fisher, MD (Glasgow Royal Infirmary, Glasgow, Scotland)
The quintessentially Scottish Dr. Fisher gave a skillful overview of SGLT-2 safety data, with particular emphasis on currently ongoing long-term outcomes studies for the drug class.
- SGLT-2 inhibitors lower blood glucose by eliminating the re-uptake of glucose in the kidney. AZ’s Forxiga/Farxiga (dapagliflozin) and J&J’s Invokana (canagliflozin) are approved and available in both Europe and the USA. Invokana launched in the UK in late February, just before the conference.
- SGLT-2 inhibitors have a number of effects that might positively impact cardiovascular safety. These include: 1) reducing A1c; 2) possible weight reduction; 3) reductions in blood pressure; 4) no effect on heart rate; and 5) lipid changes (increase in HDL and a decrease in triglycerides, although total cholesterol and LDL rise).
- In two meta-analyses, dapagliflozin was not associated with an increase in cardiovascular disease. In fact hazard ratios were lower than 1.0 and upper 95% confidence intervals were less than 1.3, so there is “a hint of a reduction” in cardiovascular disease, in the words of Dr. Fisher. Certainly, the Kaplan-Meier plots showed consistently lower event rates over time with dapagliflozin.
- Accordingly, DECLARE-TIMI58 is a recently initiated study of dapagliflozin intending to enroll 17,150 patients (for 4.5 years) and designed to prove cardiovascular superiority – i.e. that dapagliflozin lowers cardiovascular risk (composite MACE) in patients with established cardiovascular disease or multiple risk factors.
- Analysis of the Phase 2/3 study data for canagliflozin through November 2012 again shows no increase in overall cardiovascular risk, with a hazard ratio of 0.9 and an upper bound of the 95% confidence interval of 1.2. There was a hazard ratio for stroke above 1.0 that merits further attention, but it was not statistically significant.
- The CANVAS cardiovascular outcomes trial for canagliflozin was set up pre-registration and has enrolled over 4,000 people. Some unblinded data from CANVAS was included in the submissions to the FDA, but the trial will continue on. CANVAS R is an additional study on renal endpoints (but will also track MACE). CREDENCE will assess both renal and cardiovascular endpoints in patients with diabetic nephropathy, and is looking to investigate whether canagliflozin might have a glucose-independent benefit for patients with nephropathy (read our report).
Questions and Answers
Q: To what extent can we generalize from the patients in the high-risk trials to the average patients?
A: This is a poignant question. Data suggests that it is extremely difficult to generalize over long periods of time. Many trials have shown that incidence of MACE events is actually worse in the short term when there might be benefits in the long term. So I share your ‘don’t know’ on this one. However, it can help in some clinical use, because many people are diagnosed at the time of a coronary event. So it is at least useful to know how to deal with these people.
Q: What is the role of the key opinion leader when it comes to impressions of safety? We are now moving to an area where lots of these results are pretty difficult to interpret.
A: The trouble with opinion leaders is that you get ten different opinions!
Q: As we uncover specific risks related to drugs, how do we give patients informed choices?
A: We look at data from large groups and try and apply it to an individual.
Q: What do you think about the appropriateness of doing interim analyses?
A: Actually it’s the FDA that has asked for interim analyses. The issue is that if you reveal the interim data, then the investigators or participants may behave differently in the rest of the trial.
A: That’s an extremely difficult one. You want to see the endpoint of the trial. We’ve seen several massive trials broken into early and so we don’t know the long-term effect. We should be prepared to accept post hoc analysis.
David Russell-Jones, MBBS, PhD (University of Surrey, Guildford, UK)
Dr. Russell-Jones gave a most valuable presentation of new insulins, including Novo Nordisk’s Tresiba (insulin degludec), Sanofi’s U300 insulin glargine, Lilly’s Peglispro (Bil; LY2605541), Lilly/BI’s biosimilar insulin glargine, and Novo Nordisk’s FIAsp (NN 1218). He was clearly enthusiastic about the pace of development and optimistic that the new options would be helpful for patients.
- It’s been shown (in the 1970’s) that the effect of insulin at normal concentrations is to suppress glucose output from the liver; significant glucose storage in muscle only occurs at higher concentrations. That physiology is why beta cells are in the pancreas, and release insulin into a portal system that goes directly to the liver. Subcutaneous administration of insulin tends to over-insulinize the muscle and under-insulinize the liver. As Dr. Russell-Jones pointed out, not only does insulin have the wrong PK/PD profile, but we are also giving it in the wrong place.
- Novo Nordisk’s Tresiba (insulin degludec) has a very long duration of action, with a half-life of 25 hours, gradually reaching a completely flat profile after six days of once-daily administration, and allowing for flexible dosing. Tresiba is currently available in Europe (see our Novo Nordisk 4Q13 Report for notes on its performance post-launch). It has a “neat” protein structure with an added glutamic acid spacer and a fatty acid side chain, which polymerizes the insulin under the skin. This polymer slowly dissolves, leading to long duration of action. There is an extensive set of trials (BEGIN) that show non-inferiority to Sanofi’s Lantus (insulin glargine). However, the nocturnal hypoglycemia event rate is lower with Tresiba than Lantus. Tresiba’s flat profile allows for “wacky” flexible dosing (e.g. doses 40 hours then 8 hours apart) – so patients don’t have to feel guilty if they are a little bit late taking their dose.
- Sanofi’s U300 insulin glargine is a concentrated formulation of its bestseller Lantus, but offers flatter kinetics and a slightly longer duration of action. It is the same molecule as glargine, but in a lower volume, so the crystals are a little closer packed, changing the kinetics. Accordingly, U300 has a flatter kinetic profile and seems to last longer. The EDITION trials showed no difference in A1c-lowering efficacy to traditional U100 insulin glargine, but with lower nocturnal hypoglycemia – read our report on some of the more recent EDITION data. U300 insulin glargine is in very late phase 3, and should be with us in a year or so.
- Lilly’s peglispro (Bil; LY2605541), is a significantly hepato-selective (liver-selective) insulin. Peglispro has a very long half-life (24-48 hours), possible weight benefits, and a suggestion of better efficacy than Sanofi’s Lantus. In peglispro, a large (20kDa) PEG chain has been added to the insulin, giving the appearance of a much bigger molecule. Its hepatoselectivity means it has a much greater effect on suppressing glucose from the liver, but a lower effect on reducing peripheral glucose uptake – which could potentially be more physiological. In a treat-to-target crossover phase 2 study, the data is suggestive that A1c might be lower than glargine, but we will need to wait for phase 3 data to be certain. Peglispro was also associated with lower nocturnal hypoglycemia. Another, very interesting aspect is that being hepato-selective yields weight loss for patients originally taking Lantus. There are two concerns: in type 1 patients, the lipid profile worsened slightly, which may or may not have clinical significance – we will only know from the larger trials. There is also a slight rise in liver enzymes in both type 1 and type 2 diabetes patients. The phase 3 program is large, with over 6,000 patients. Results will start coming out at end of 2014 or the beginning of 2015.
- Lilly/BI’s biosimilar glargine – to all intents and purposes should be a very similar molecule to Sanofi’s Lantus, but may have an effect on the market if it’s cheaper. This is the first of many biosimilar insulins that are coming (Merck is developing one with the Korean biosimilar manufacturer Samsung Bioepis, and Biocon has a biosimilar insulin glargine available in limited markets).
- FIAsp (NN 1218) is a faster-acting variation of Novo Nordisk’s NovoLog/NovoRapid (insulin aspart). Clinical trials are taking place (Onset program), but there is little data available currently. Post-meal excursions appear better with FIAsp.
- Lilly’s Humalog U200 – no data has been published to date.
- Fixed dose combinations of insulin and GLP-1 agonists will be very interesting. These include Novo Nordisk’s IDegLira (insulin degludec and liraglutide) and Sanofi’s “Lixi/Lan” (insulin glargine and lixisenatide).
- There are a number of oral insulin candidates in development, from Novo Nordisk, Oramed, and Biocon. They use different approaches for delivering insulin through the GI tract and into circulation, and mostly focus on long acting insulin (see our report from late 2013 on Novo Nordisk’s oral insulin portfolio).
Questions and Answers
Q: Are there any subtle differences with biosimilar glargine?
A: It’s going to made in a different way in a different place, so we expect that it won’t be 100% identical. For whatever reason, patients might believe there to be a difference, and we might have to address that.
Q: Are there any studies on liver steatosis with peglispro?
A: There are animal studies only. I would like to investigate this, but we can’t get the insulin until the Phase 3 program is over.
Q: Won’t the presence of U100, U200, and U300 insulins be very confusing?
A: Hospitals are extraordinarily dangerous places for our patients. Mistakes are made all the time. We will have to tighten up.
Targeting the Kidney to Treat Diabetes: SGLT-2 Inhibitors
John Wilding, DM (University of Liverpool, Liverpool, UK)
In this presentation, Dr. John Wilding bombarded us with vast amounts of clinical trial data relating to J&J’s Invokana (canagliflozin) and AZ’s Farxiga/Forxiga (dapagliflozin). He appeared enthusiastic about the new class of SGLT-2 inhibitors, and stated that they will be valuable for patients due to their good glucose control, weight loss, and utility as an add-on therapy. Their side effect profile (genital and urinary tract infections) is manageable and otherwise they appear safe and might even turn out to be cardioprotective.
- The kidney filters and reabsorbs up to 180g of glucose per day from/to the blood. This happens via the action of sodium-glucose co-transporters SGLT-1 and SGLT-2. In the kidney, SGLT-2 is responsible for 90% of the glucose reuptake. The remaining 10% is handled by SGLT-1. In a healthy person, at around 150 mg/dl, filtered glucose begins to be excreted in the urine and by 240 mg/dl, the rate of glucose reabsorption has reached a static maximum – any excess appears in the urine. Interestingly, in diabetes, glucose reuptake capacity actually increases (a maladaptive phenomenon).
- SGLT-2 inhibitors prevent glucose re-uptake, leading to lower average glucose along with possible weight loss and blood pressure improvements. Early studies show the elimination of up to 60g of glucose per day. Given the mechanism, concerns with the drug class would be an increase in urinary tract and genital infections, hypoglycemia, impact on electrolytes and renal function, and cardiovascular safety
- Dr. Wilding presented a bevy of clinical data for dapagliflozin and canagliflozin:
- In clinical trials, dapagliflozin showed a 0.8% A1c reduction after two years versus placebo, and weight loss was around 3 kg. In head-to-head studies against a sulfonylurea, dapagliflozin showed A1c superiority and a big weight improvement sustained out to four years. Very similar data was obtained with canagliflozin.
- Canagliflozin showed A1c superiority against Merck’s DPP-4 inhibitor Januvia (sitagliptin) after a year (at the 300mg dose).
- Dapagliflozin added on to insulin showed a 0.5% A1c improvement compared to placebo and a 3 kg improvement in weight. There was also no change in daily insulin dose with dapagliflozin, while the placebo group increased by 18U over the two year period.
- Across the entire spectrum of usage in type 2 diabetes (mono-, dual-, triple-therapy and add-on to insulin) we see a class reduction in A1c (of ~1% with orals), roughly a ~3kg weight loss and a ~4 mmHg drop in systolic blood pressure. In a body composition study using DEXA, the majority of the weight lost was determined to be fat. Blood pressure improvement is because of a mild diuretic effect.
- Genital and urinary tract infections affect 1%-10% of patients depending on the trial, but other side effects are very uncommon. Genital infections in women tend to occur within the first six months in 10-12% of patients; they tend to occur once and respond to treatment. Men behave similarly, but have lower rates. There are no significant effects on bone mineral density, only minor changes in electrolytes and in the long term, there is no effect on kidney function (eGFR).
- Although we see some small negative changes to lipids, the best data currently available do not show increased cardiovascular risk with either dapagliflozin or canagliflozin. There was even a hint that they might be cardio-protective, but we will have to wait for the results of long-term outcome studies to be certain.
Questions and Answers
Q: Is there any connection between dapagliflozin and bladder cancer?
A: No. All the data were reviewed carefully by the FDA. Even though there was excess bladder cancer with dapagliflozin, it was ruled out as causative. The bladder cancers were spotted early into the trial and most patients presented with hematuria at baseline. It normally takes a very long time for an agent to cause cancer, so it’s really likely that the cancers were pre-existing.
Current and Upcoming Cardiovascular Disease Trials In Diabetes: What Do They Tell Us?
John Petrie, MD, PhD (University of Glasgow, Glasgow, Scotland)
Dr. John Petrie gave us an overview of all the large cardiovascular outcomes trials in type 2 diabetes that are expected to report over the next few years. He welcomed the trials, since he relished the value of all the patient data. However, he noted that there are some signs that the FDA is feeling that it might have overreacted in the aftermath of the rosiglitazone affair and that the volume of trials may not continue into the future. Obviously it would be a big deal for the FDA to move directions on this front.
- The current situation with cardiovascular outcomes trials goes back to the rosiglitazone affair. Rosiglitazone (GSK’s Avandia) was taken off the market in the EU and severely restricted in the USA, despite the manufacturer undertaking a large scale cardiovascular outcomes trial. One of the issues was that the RECORD trial was open label and was criticized but then vindicated, and it seems that the drug should never have been restricted on cardiovascular grounds (see our report on the 2013 Advisory Committee on the re-adjudication of RECORD).
- DPP-4 inhibitors – The EXAMINE trial for Takeda’s Nesina/Vipidia (alogliptin) showed that alogliptin has no cardiovascular risk, with a hazard ratio (HR) of 0.96 (CI 1.16) compared to placebo (the margin of safety specified by the FDA is 1.3 for the upper confidence interval). Equally, SAVOR (TIMI-53) for AZ’s Onglyza (saxagliptin) showed a HR of 1.0 for saxagliptin versus placebo. However, there seems to be a higher hospitalization rate for heart failure in both trials (statistically significant only in SAVOR), which has not been explained. It appears that markers for heart failure were higher in the saxagliptin arm of SAVOR at baseline, so the jury is firmly out. The latest assessment by the FDA and EMA on pancreatitis and pancreatic cancer, which was very recently completed (read our report), indicates that they are not seeing a worrying signal in current data. Trials yet to report include TECOS (Merck’s Januvia [sitagliptin]) and CAROLINA/CARMELINA (Lilly/BI’s Tradjenta/Trajenta [linagliptin]).
- GLP-1 Agonists – The key trials under way are ELIXA (for Sanofi’s Lyxumia [lixisenatide], reporting in January 2015), LEADER (for Novo Nordisk’s Victoza [liraglutide], reporting in October 2015), EXSCEL (for AZ’s Bydureon [exenatide once weekly], reporting in April 2018), and REWIND (for Lilly’s dulaglutide, reporting in April 2019). There also studies planned with GSK’s Eperzan (albiglutide).
- SGLT-2 inhibitors – The key trials underway are BI 10773 (for BI/Lilly’s empagliflozin – unfortunately the FDA have stopped the approval process this week because of manufacturing issues, which Dr. Petrie expressed could “hopefully” be resolved soon), CANVAS (for J&J’s Invokana [canagliflozin], reporting in June 2018), and DECLARE (for AZ’s Forxiga/Farxiga [dapagliflozin], reporting in April 2019).
Dr. Petrie welcomed the trend in performing these major trials, but noted that it may not be maintained. There are some signs that the FDA thinks that it could have over-reacted and might step back, he said. CVOTs have generated lots of data, but it could be argued that the resources might have been better spent on patient care. Nonetheless, they incorporate 200,000 patients, will provide valuable learning, and the results could even change the order in which therapies are used.
Corporate Symposium: Individualization Of Patient Care Versus National Guidelines: Is There A Conflict? (Sponsored by Takeda)
This House Believes That All Patients Should be Treated According to National Guidelines
Sarah Jarvis, BMBCh (Richford Gate Medical Practice), Azhar Farooqi, OBE (Leicester City Clinical Commissioning Group, Leicester, UK)
This lighthearted seminar was designed as a “mock debate.” Dr. Sarah Jarvis, did an entertaining turn in “supporting” the motion – making the (serious) point that there is a four-fold variation in the quality of diabetes care across the UK, but mostly sounding like she wanted everyone to slavishly follow the rules. For his part, Dr. Farooqi invoked dark conspiracy theories with his talk of the “guidelines industry.” Ultimately this was just a fun smokescreen to make the case for individualized treatment. In any case, everyone accepted that ‘guidelines’ by definition provided only guidance, that NICE (the UK’s clinical standards body) specifically included “wiggle room” within its guidance, and that the trend was towards the ADA/EASD’s more explicit recommendation that no one size fits all. In a follow-up presentation, Dr. Jarvis discussed the main advantages of Takeda’s DPP-4 inhibitor Vipidia (alogliptin; branded Nesina in the US). She noted that Vipidia was the only DPP-4 to have a completed cardiovascular outcomes trial with no safety concerns (the SAVOR CVOT for AZ’s Onglyza [saxagliptin] found a statistically significant increase in heart failure), that it was proven superior to sulfonylureas after two years when used with metformin, and that it was 16%-20% cheaper than other DPP-4 inhibitors in the UK, where Vipidia and Vipdomet (metformin FDC) were launched recently. She asserted that on average, Vipidia saves £84 ($140) per patient year. During Q&A there was an interesting discussion in which the audience agreed that patients are generally accepting of switching to lower cost therapies if they are told that they are equally effective. The British National Health Service (NHS) is increasingly making doctors accountable for reducing costs.
Questions and Answers
Q: How does this compare to Lilly and BI’s Trajenta/Tradjenta [linagliptin] in patients with kidney disease?
A: Well, in patients with CKD, you have to change the dose [of alogliptin], but linagliptin at CKD stage 4 is basically monotherapy – you have to take away sulfonylureas and metformin and it’s not licensed for use with a TZD. DPP-4 inhibitors are really for early stage use, not for patients near end stage renal disease, who probably need insulin.
Q: It’s important for us equally to know when to stop prescribing these drugs, since they are very expensive.
A: I couldn’t agree more. However, bear in mind, although sulfonylureas are cheap, test strips and hypos are not.
Banting Memorial Lecture
Technology And Diabetes Care: Appropriate And Personalised
John Pickup, MD (King’s College London School of Medicine, Guy’s Hospital, London, UK)
The legendary Dr. John Pickup began this prestigious Banting Lecture with the sentence, “It may surprise you to know that I actually hate technology… from “TwitFace” [social media], to the “unexpected husband in the bagging area” [supermarket self-checkout]. Thus setting a humorous tone, he proceeded to take us on a tour of the past, present and future of diabetes technology, taking in the first portable closed loop system (1963) to the latest in nano-encapsulation of islets. He explained why meta-analyses of pumps and CGM don’t show enough value (they don’t include the worse patients) and showed us three future uses of nano-technology (smart tattoos for glucose measurement, nano-encapsulated islets and oral nano-encapsulated insulin). For Dr. Pickup diabetes technology must be targeted, personalized and appropriate. And it must make an effective and affordable contribution to diabetes care.
- The birth of diabetes technology stems from the mid-1960s to the mid-1980s. In this period we had the portable closed loop (1963), the first portable blood glucose meter (1968), the insulin pump (1976), the insulin pen (1981), CGM (1982) and electrochemical SMBG (1984).
- Modern diabetes technology has clear effectiveness but some contradictions. For example:
- Some trials show effectiveness, some do not;
- Equal access to healthcare is critical, but there is highly variable availability of the technology;
- Technology can be expensive, but in the long term it might be more cost-effective;
- Patients are overwhelmingly positive, but many healthcare professionals are still negative.
- Illustrating the above, Dr. Pickup demonstrated how meta-analyses of pump and CGM effectiveness need to take baseline A1c into account. In pumps, Dr. Pickup’s meta-analysis showed that A1c is lowered by 0.6% and hypoglycemia reduced 75% compared to MDI, but other meta-analyses have shown no significant value. Often, patients are excluded if they are badly controlled (but these are the ones with most to gain from pumps), or timeframes are not long enough. In both pumps and CGM, there is a strong effect of baseline A1c – for example those starting with A1c of 10% on MDI can get to 8% on the pump.
- Dr. Pickup’s (rather thrifty) algorithm for diabetes devices would be to take MDI patients and give them full support. Pumps would be suitable for the 20% of patients failing this regimen, and CGM appropriate for the 20-30% having problems with the pumps.
- According to Dr. Pickup, the roadmap for “feedback controlled insulin delivery” is likely to be:
- 1. Low glucose suspend (which can cause a 96% reduction in hypoglycemia for the top quartile, or completely eliminate severe hypoglycemia in problem patients)
- 2. Predictive low-glucose suspend
- 3. Predictive LGS plus automated insulin delivery for hyperglycemia
- 4. Fully closed loop overnight only
- 5. Fully closed loop night and day
- In type 2 diabetes, pumps work well for those with poor control on MDI. Patch pumps might be a good future alternative in type 2. Patch pumps are smaller and easier to use with simpler infusion regimens, they have a simple manual bolus system, are usually attached to the body with adhesive (hence the name) and ‘tubeless’ in design.
- Looking to the (nano)future (1): Smart tattoos for measuring glucose. A fluorescent-labeled bacterial glucose binding protein could be implanted subcutaneously or intradermally and illuminated with near-infrared light to obtain the glucose concentration. The tattoo would otherwise be invisible. Fluorescence is a great way to measure glucose, since it is very sensitive, free from interference, and the signal is not affected by coating or degradation in vivo.
- Looking to the (nano)future (2): Nano-encapsulation of pancreatic islets. Islets could be “shrink wrapped” to keep out immune cells while still allowing oxygen, glucose and nutrients to enter. Islet coatings are built layer by layer. In animal studies, non-coated islets lasted 7-10 days, compared to 28-37 days before coated islets were deliberately removed.
- Looking to the (nano)future (3): Oral encapsulated insulin. Insulin coated sequentially in chitosan/PEG and heparin/PEG can survive the harsh environment of the stomach. In mouse studies, nano-encapsulation lowered glucose more than uncoated insulin.
- Future diabetes technology must fit two criteria: it must be targeted, personalized and appropriate. And it must make an effective and affordable contribution to diabetes care. If that’s the case, then we can learn to like it!
Diabetes Technology: Closed Loop And Pump Audit
Closed Loop Results
Roman Hovorka, PhD (University of Cambridge, UK)
The legendary Dr. Roman Hovorka presented results from his overnight closed loop trials in adolescents. We’ve seen much of this data before, most recently at ATTD (see our Day #3-4 Report), but still can’t help but be impressed. In particular the closed loop algorithm increases time in target, improves mean glucose, decreases severe hypoglycemia, reduces daily insulin needs, and takes a large mental burden off the patient. Dr. Hovorka is currently applying for regulatory approval for a three-month continuous wear outpatient trial, which is particularly exciting.
- Dr. Hovorka presented his data from overnight closed loop home studies in adolescents with type 1 diabetes, using the “Florence” prototype equipment. The Florence prototype incorporates the Abbott Navigator CGM, the Companion receiver, a Dana R insulin pump, and a small “palmtop” computer that runs the control algorithm. This MPC algorithm modifies insulin delivery every 12 minutes and is initialized by body weight, total daily dose and basal pump settings. The algorithm aims for a target range of 70-144 mg/dl.
- Patients performed 21 consecutive nights of closed loop control and 21 nights of CGM only, in a crossover design with a 2-3 week washout period. The patients were 12-18 years old with A1c <10%, CGM naïve and with more than three months pump experience. They received training for a two-week period prior to the trial and training was also performed at home for the first night. There was no remote monitoring, but there was a 24-hour support line in case of problems.
- After a total of 269 closed loop nights (80% of actual nights), time in target zone was 68%, compared to 46% with CGM only. Severe hypoglycemia (<63 mg/dl) declined from 17% to 10% of nights. There was no change in hypoglycemia (<70 mg/dl) but the baseline was very low. Average glucose was reduced from 151 to 140 mg/dl, while total daily dose of insulin was reduced from 53.2U to 49.9U. The net effect is to greatly reduce the variability of glucose, particularly towards the end of the night/early morning, and that was obtained with a big increase in the variability of insulin delivery.
- Each subject was different and each night was different. Analysis of “controller effort” - defined as the percentage of insulin provided by the closed loop compared to the control basal program - showed broad inter- and intra-patient (night to night) variability.
- A three month home closed loop trial has been submitted to regulatory agencies. The new system consists of the FreeStyle Navigator (Abbott) sensor and receiver, a Dana pump, and an Android phone (carrying out the communication and control).
Questions and Answers
Q: How far ahead does the algorithm look?
A: It looks 2.5-3 hours ahead; it also takes into account the absorption of meals and prandial boluses.
Q: On one chart you showed, it took four hours to reach the target - that seems like a long time.
A: It takes about four hours on average to go from a wide range to a tight range, depending on starting glucose, and because we want to be very careful not to overdose on insulin. Sometimes it takes longer – we have had some patients take 20 hours, to settle down. But in terms of going faster, remember that the closed loop can’t do anything that you couldn’t do yourself manually.
Psychosocial Aspects Of Closed Loop
Katharine Barnard, CPsychol (University of Southampton, Southampton, UK)
Dr. Katharine Barnard had some very interesting things to say about patient and parent attitudes in overnight closed loop studies. Ultimately, the goal of these trials is improved quality of life, so it makes sense to understand how participants are feeling. The results of semi-structured 30-minute interviews of patients in Dr. Roman Hovorka’s adolescent overnight closed loop studies were striking. Patients felt safe, they felt great about the clinical benefits, and felt that they were getting better sleep. Parents appreciated the reassurance in both the short and long term, and also not having to get up in the night to test their child. On the negative side, there were some issues raised with the reliability and size of the device, as well as alarm fatigue. We imagine that this latter set of problems will be overcome in future generations of the technology.
- The ultimate goal of any therapy is an improvement in quality of life, over and above the clinical benefits, so it’s important to ensure that the closed loop can improve lives and that patients can deal with it. Living with diabetes is mentally very burdensome – some people can fit it into their lives very well, yet others can’t. Equally, closed loop trials demand a lot from participants and they raise large expectations. Therefore, it is important to understand and support the psychosocial impact of starting the closed loop and discontinuation following participation.
- Dr. Barnard personally conducted 30 minute semi-structured interviews with the adolescent participants in Dr. Roman Hovorka’s overnight closed-loop studies, to explore the positive and negative aspects of the experience. She conducted fifteen interviews, and also interviewed the parents of the participants.
- The key messages from adolescents (positive and negative) were: “felt safe,” “stable blood glucose levels,” “better sleep,” “calibration difficulties,” “discomfort/size of device,” and “alarms beeping.” Having trust in the device was considered significant. On the negative side, many of the issues were due to technology and design factors that continue to be improved – particularly size. However, the Dana pump was not generally popular.
- The key messages from parents were: “reassurance/peace of mind,” “improved blood glucose control,” “equipment/useful data,” and “better start to day.” Typically, parents didn’t get up to test blood in the night, which was a huge change for them. They also had a large weight off of their shoulders because they trusted that the device would work. They also believed that starting the day with a good blood glucose led to better control well into the day. The potential that this could alleviate the risk of long-term complications was also significant. The negatives were similar to the childrens’: “technical difficulties,” “alarms beeping,” and “too big.”
- Nobody regretted taking part in the trial, but participants were often frustrated by the technology.
Questions and Answers
Q: Should we profile patients before studies?
A: In all technology trials, is it possible to identify participants that will be most suited to the technology up front? I genuinely don’t know if you could profile those people that would be most suited to participate in the trial.
Q: Why were patients thinking they were better during the day, since it only ran at night?
A: Because when patients woke up on a good number, they thought they had better control up through lunchtime.
National Insulin Pump Audit: Adult Data
Philip Weston, MD (Royal Liverpool University Hospital, Liverpool, UK)
Dr. Philip Weston presented the results from the first UK-wide pump audit. It seems that getting a pump is not too difficult if a patient meets the NICE criteria, and most current patients are using pumps as recommend by NICE. The audit suggests significant clinical benefits of pumping, and once on the pump almost all patients continue.
- The British National Health Service (NHS) funded an insulin pump audit, which has just been completed. It’s not certain whether funding will be forthcoming for future audits. The audit had two pieces: a service level audit (numbers of patients, funding, and education), and a patient level audit (demographics, indications, outcomes). The results presented today were for the patient audit; Dr. Weston presented data for adults, while Dr. Atrayee Ghatak presented pediatric data (see below).
- There are n=13,428 adults on pumps in the UK, based on returns from 97% of the 183 sites that initiate pump therapy. This represents about 6% of type 1 adults (221,000). NICE (National Institute for Health and Care Excellence) estimate that around 12% of adult type 1 patients will meet criteria for pumps, so there is some opportunity to increase penetration. According to Dr. Weston, the UK is “falling woefully behind” other European countries and especially America.
- The key results from the national insulin pump audit were:
- Most centers rely on pump companies to train patients. There is a shortage of qualified diabetes staff.
- There were relatively few pump funding issues. Pump starts have risen significantly in the last four years. In the UK there are about 1,000 new pump starts per year. The reasons for starting on a pump were poor glycemic control (51%), hypoglycemia (37%), hypoglycemia/hyperglycemia unawareness (21%), other (21%). Most of those obtaining pumps fitted the NICE criteria.
- Pumpers are 68% female, and 98% continue pump therapy.
- Data suggests that pump therapy lowered A1c by about 0.4% in type 1 patients and by about 1.5% in type 2 patients (did not achieve statistical significance).
- The frequency of hypoglycemia is greatly reduced on the pump, although the overall level is still too high.
- Rates of hospital admissions for DKA were similar to the general population.
National Insulin Pump Audit: Pediatric Data
Atrayee Ghatak, MD (Alder Hey Children’s Hospital, Liverpool, UK)
In the presentation of pediatric data from the national insulin pump audit, Dr. Ghatak showed that children are not generally having problems getting pumps in the UK, that the NICE guidance is only loosely followed, that the audit suggests significant clinical benefits from pumping (A1c, hypoglycemia frequency, hypoglycemia awareness), and that 70% of centers ignore the NICE advice to give pumpers over 12 a “trial” of MDI.
- This presentation summarizes the pediatric data from the national insulin pump audit.
- The NICE guidance for children is for kids <12 years, pumps are recommended if MDI is thought to be impractical or inappropriate; for kids >12 years, pumps are recommended for patients with disabling hypoglycemia or A1c >8.5%. A really important caveat is that NICE also recommend giving pediatric pumpers a trial of MDI once they are over 12, although 70% of centers don’t do this.
- The key results of the pediatric audit are:
- There are estimated to be roughly 10,000 children on pumps in the UK.
- The peak in pump starts is around ages 11-14, which is the peak age for diagnosis of diabetes. 63% of starts are below the age of 12 years, and 47% of children with diabetes are under 12, implying that pumps are being offered to younger and younger patients. Roughly 60% of kids started pumping within the first three years of diagnosis.
- Funding is generally becoming more available for pump services
- The indications for starting pumps are “poor glycemic control,” “troublesome hypoglycemia,” “hypoglycemia with loss of awareness,” and “other,” which includes age, quality of life, patient choice, parental choice, needle phobia, and sports. The NICE guidance is only “loosely followed.”
- A1c on the pump decreased from 8.7% to 8.4% and 43% of children showed a decrease in hypoglycemia events.
- Hospital admissions for DKA and hypoglycemia were similar to the diabetes population at large. The rate of DKA is slightly higher for pumpers, but they were supposed to be the less well-controlled patients.
- 98% of children continued on pumps
Questions and Answers
Q: I am interested in pump/infusion set data. Did you collect any information about failure rates? EASD is preparing a statement about reliability of medical devices and post market surveillance, and this might be interesting.
A: We asked which companies’ pumps are being used, but we haven’t looked at that data in detail. But we have no other information on pump failure.
Q: For A1c it wouldn’t it make more sense to stratify by starting A1c?
A: We know that those that start highest fall the most; we will discuss this in more detail in the publication.
Q: What about the centers that don’t provide pump therapy?
A: We haven’t looked at centers that don’t provide pump therapy. What are they doing? My concern is that patients would benefit from pumps, but are being steered away by the clinical teams.
Symposium: The Next Frontier in Diabetes Management – Abbott Diabetes Care
Iain Cranston, MD (Portsmouth Hospitals NHS Trust, Portsmouth, England), Gerry Rayman, MD (Ipswich Hospital NHS Trust, Ipswich, UK), Anita Bowes (Royal Bournemouth Hospital, Bournemouth, England), Jared Watkin (VP of Technical Operations, Abbott Diabetes Care, Alameda, CA)
This symposium, presented by Abbott Diabetes Care, covered Abbott’s new flash glucose monitoring product, which has a fourteen day wear and doesn’t need fingerstick calibration. The session was similar to sessions we’ve reported on from EASD and ATTD – the company has been doing an effective job at getting the word out. We could not be more excited by the new product, which Abbott hopes will be approved in the EU in 2014. Today’s symposium was well attended, and was notable for a candid Q&A session in which Dr. Jared Watkin discussed some of the product details. It seems clear that Abbott want to access a broader market than CGM has been able to tap, in particular type 2 patients. The company is in the “early stages of clinical trials,” and Dr. Watkin discussed with us plan for three major trials – one pivotal for EU approval, and two testing superiority to BGM
- Dr. Iain Cranston noted that the systematic use of SMBG or CGM data can have a powerful effect on glucose control. However both technologies have drawbacks – SMBG is potentially painful, individual measurements can be biased, and data analysis is often idiosyncratic. On the CGM side, there are problems with device accuracy, consistency, calibration, and false alarms. Data interpretation can be difficult because of the density of the data and frequent therapy changes by the patient. Last (but not least), CGM suffers from relatively poor reimbursement.
- Dr. Gerry Rayman discussed AGP (ambulatory glucose profiling). AGP takes fourteen days of blood glucose data (for predictability) and creates a modal day profile, with a line for median glucose and shaded regions for the 25th/75th percentile and the 10th/90th percentile. Since the AGP shows the variation in glucose, it is more useful than A1c in assessing the quality of glucose control. Studies show that patients find the AGP easy to understand, can see what actions need to be taken, and feel more confident in making treatment decisions based on the data. Dr. Rayman believes that AGP is superior to a classic “spaghetti” modal day report.
- Ms. Anita Bowes presented case studies in type 1 diabetes, based on AGP. In each case, she demonstrated an approach for dealing with hypoglycemia if present, reducing glycemic variability (made obvious by AGP), and then lowering the overall median glucose. A test panel of diabetes experts assessed twenty case studies with or without AGP, and concluded that it helped them identify key trouble spots with speed and ease.
- Abbott’s Mr. Jared Watkin delivered his much-appreciated presentation on Flash glucose monitoring – see our ATTD Day #1 Report for his most recent previous presentation on the topic. The details of this presentation remained largely the same: Abbott’s forthcoming product eliminates fingerstick calibration and uses a sensor about the size of a two Euro coin with a fourteen day wear time. The patient simply waves the receiver close to the sensor to obtain up to eight hours of stored continuous glucose data. The system works through clothing and the sensor is water resistant. The receiver reports time in target as well as current score and arrows. The patient’s AGP is available when the data is uploaded to a computer. As mentioned at ATTD, the latest sensor study gave a retrospectively calibrated MARD of around 10%, which is what Dr. Watkin is hoping for when development is complete. Abbott hopes to achieve CE marking in the second half of 2014, and has not yet announced plans for submission in the USA. Abbott is currently conducting early-stage clinical trials on the product now. Dr. Watkin told us that Abbott is currently designing three key trials: one pivotal for EU approval, and two to show superiority to BGM for type 1 and type 2 diabetes. Abbott believes that the product will appeal to a fairly broad market. The product doesn’t yet have a name, and the official line is that “Abbott is designing a concept for the Flash Glucose Monitoring category as an alternative to BGM for a broad population of people with diabetes.”
Questions and Answers
Q: What will be the cost of the sensors?
A: We are aware that it’s a challenge for reimbursement. We are not able to comment on cost at this point, but we do understand the issue.
Q: How does it talk to the receiver?
A: It uses RFID – a widely used wireless communication protocol. You have to get it fairly close, but it works through clothing.
Q: How does it fit in with an active lifestyle?
A: Well it’s water resistant, which means tolerant of bathing, showering and swimming, It’s not designed for scuba diving, but it’s OK for most activities.
Q: How does the sensor signal decay over 14 days? Is there an increase in variability?
A: We see no decay in response over the 14 days of wear. That’s one of the benefits of the technology. You need this to go fingerstick free, since calibration is designed to reset sensor sensitivity over the period.
Q: What is the size of sensor under the skin?
A: The sensor is about the thickness of a hair, and it goes about 5mm vertically into the skin. It’s fantastically small - about the size of two-pound coin, a little thicker and it’s really light. [Note: there was a person in the audience familiar with the product that told us that there is also an insertion device].
Q: Is there any possibility that when receiver is in range it could alert you if you were high or low?
A: Regarding alarms, we are not designing flash glucose monitoring as an alternative to CGM. We are designing it to be an alternative to BGM for a larger population. You are unlikely to get anything you could treat as a hypoglycemia alarm.
A: But it’s very convenient compared to BGM – you don’t have to wake up your kid in the middle of the night to get a reading, or a truck driver can test in a second.
Q: Do patients tolerate the sensor for 14 days? Have there been any local site reactions?
A: We are the early stages of clinical trials. The pivotal trials are initializing now. There is no indication of any significant site reactions so far over the fourteen days. It is designed to last fourteen days, and we don’t see any signs of reaction, but as we go through the pivotal trial we will be checking on that.
Q: It’s interstitial glucose, not blood glucose that is being measured. Is there a time delay?
A: Yes, there is a lag between interstitial glucose and blood glucose. The physiological lag is perhaps ten or fifteen minutes. Our experience is that patients get used to it. It gets harder when blood glucose is changing rapidly, but our device will tell you that.
Q: Can you tag medication in the handheld?
A: Yes, the reader allows you to tag medication, food, exercise, with a timestamp on the device. You can enter carbs as well.
Q: Can you tell me more about the technology? I am having difficulty imagining something under the skin for 14 days and staying the same, please explain.
A: You need good biocompatibility and stable chemistry. We’ve done a lot of work on membrane selection, and we have worked on the chemistry, to make sure it’s stable. Beyond that I can’t say much more.
Animas Press Briefing - Blood Glucose Battle: Managing Childhood Diabetes
Blood Glucose Battle: Managing Childhood Diabetes
Caroline Chivers (Marketing Manager, Animas UK), Ramakrishna Venugopalan, PhD (Director Artificial Pancreas Initiative, Animas, Philadelphia, PA, USA), Fiona Campbell, MD (St. James’ University Hospital Leeds, Leeds, UK), Karen Baxter (Country Manager UK & Ireland, Animas)
Animas mounted a presentation for press and advocates, celebrating the recent pediatric approval for the Animas Vibe.
- The Animas Vibe has now been approved for use together with the Dexcom G4 Platinum CGM in children 2-17.
- The pediatric indication is very important for Dr. Fiona Campbell, who is a leading user of diabetes technology in children. She believes that technology should be integrated with routine care. Of her 480 pediatric patients, 55% use pumps and have access to CGM (either permanently or intermittently). Her future wish is that she can fit a bionic pancreas to newly diagnosed kids just like giving them a mobile phone.
- In an interactive presentation, the challenges of living with diabetes for children were highlighted and the advantages of using the Animas Vibe pump were explained.
- Dr. Ramakrishna Venugopalan showed some dQ&A data suggesting that pump patients’ biggest priorities are accurate delivery, comfort, BGM/CGM integration, and overall safety. Two key areas that address these needs are Animas’ patented drive technology (384:1 drive ratio) and an accurate CGM sensor (MARD of around 11%). The philosophy is to take the best pump and CGM technology and build on it to create an artificial pancreas.
- Animas has worked with leading closed loop academics and the JDRF to develop a prototype system. Three clinical studies have been completed and presented. The focus is predictive control, rather than reactive control.
- Key next steps are to engage with the FDA and payers on transitional and pivotal clinical study design. It is not efficient to perform multiple studies, so ideally they could satisfy reimbursement as well as regulatory needs by specifically measuring outcomes.
- Dr. Venugopalan believes that Animas is a leader in closed loop because they already have commercialized a pump/CGM combination, which is the basic platform. In his view, any company that hasn’t got this platform approved is three to four years behind. There is a big difference between research studies and commercialization in terms of resource commitment. He believes that for this reason, there are only two or three companies that can bring an artificial pancreas to market.
A key issue in commercializing the artificial pancreas is addressing the interruptions of “real life” – sensor changes, set changes, sensor failures, and sensor deterioration. All these can drop the user out of closed loop. How to handle all these transitions is now a big question.
Rank Nutrition Lecture in Honor of Harry Keen
Obesity: From Genes to Behaviour
Sadaf Farooqi, PhD (Wellcome Trust-MRC Institute of Metabolic Science, Cambridge, UK)
Dr. Sadaf Farooqi delivered a fascinating lecture that served to give a sense of how complex the causes of obesity are. Her approach is to study the genetics of obesity, in the hope that it can yield some understandings of the principal mechanisms of how we maintain or gain weight. The problem is that obesity is genetically extremely heterogeneous - many genes contribute in many complex ways. Although mutations in single genes have been identified in severely obese children, they account for a tiny percentage of the cases known to date. Nonetheless, Dr. Farooqi is extremely active, studying leptin, the genetics of appetite (via fMRI), developing a cohort of 5,000 obese children (the GOOS study), creating a study of thin people (STILTS) and building an eating behavior unit at Cambridge. Although progress is slow, this is extremely important work, and it’s hoped that newer genetic techniques and a multi-disciplinary approach can bring a breakthrough.
- Over the last 30 years, there has been a big societal shift towards obesity, caused by big environmental changes against a genetic background that supports obesity. Currently in the UK, nearly 25% of the population is obese. Dr. Farooqi’s work is focused on understanding the fundamental mechanisms involved in energy homeostasis, understanding how these mechanisms can be disrupted to cause obesity, and then to use the knowledge to find therapies. Genes are a good tool for gaining this understanding. Accordingly, Dr. Farooqi has studied the genetics of 5,000 severely obese children in the GOOS study. However, all the obesity genes that have been found to date only explain 7% of the GOOS cohort, demonstrating that the genetic factors underlying obesity are highly complex.
- The brain receives signals from adipose tissue and the gut to help it maintain and adapt metabolism. An example is leptin – which regulates appetite, satiety and also which foods are desirable. fMRI shows that leptin deficient children were enthusiastic about any kind of food, but after only seven days of leptin treatment they became more discriminating.
- Mutations in the MC4R (melanocortin 4 receptor) gene appear to be one of the most common genetic forms of obesity. Mutations in the gene occur in about 1% of obese adults (a fivefold higher incidence than the allele for cystic fibrosis). MC4R is implicated in brain signaling of appetizing food. Rhythm Pharmaceuticals is currently investigating RM-493, a small-peptide melanocortin-4 receptor agonist. It yields weight loss in primates through a reduction in food intake, and is currently in Phase 1B with MC4R deficient patients.
- KSR2 is another gene, of interest that is expressed in the brain. Mice with a KSR2 variant eat more, have a lower basal metabolic rate, are severely insulin resistant, and have impaired glucose tolerance.
- The study of the genetics of obesity is ongoing, with a focus on new candidate genes and new genome-wide approaches to finding genes (such as whole exome sequencing of families). The WT UK10K consortium has a database of 1,000 UK Caucasian patients that it hopes to analyze in this way.
- Part of the approach to understanding obesity is studying very thin people, since the heritability of thinness is equal to that of severe obesity. The STILTS (Study Into Lean and Thin Subjects) has recruited 1,500 thin (BMI <18kg/m2) yet healthy people via practitioners in the UK. This segment represents about 0.2% of the population. Some can remain thin regardless of what they eat.
- Finally, at Cambridge, Dr. Farooqi is building an eating behavior unit, that will allow participants’ eating behaviors to be closely studied without giving the appearance of a lab environment.
It's Time to Stop Promoting Carbohydrates to People with Diabetes
Trudi Deakin, PhD (X-PERT Health, Hebden Bridge, UK), Carla Gianfrancesco, (Sheffield NHS Trust, Sheffield, UK)
Since the invention of insulin, dietary guidelines for people with diabetes indicated lowering carbohydrate intake and increasing fat and protein, but this changed in the 1990s along with general population guidance to eat less fat (and therefore more carbs). The focus for people with diabetes was to base the meal around ‘starchy carbs’ and consume at least 50% of calories from carbs, but this left some patients confused and with unstable blood sugars. Finally in 2010, the ADA suggested that there is no evidence that there is an ideal ratio of macronutrients for the general population. This was backed up in 2011 by Diabetes UK. The current reference intake for daily carbs is 260g for a healthy adult. The brain/blood require 130g/day, so anything under this level qualifies as a low carb diet. Ms. Deakin noted that “we need to stop promoting carbohydrates because the evidence isn’t there. We need an individualized approach.” For her part, Ms. Gianfrancesco showed that results from low carb studies are contradictory and haven’t clearly demonstrated success in managing diabetes and losing weight. However, studies are usually short term, with small numbers of participants, no control group, and often no way to be certain that participants adhered to the diets; clear evidence is very hard to come by.
- As a result, it seems that weight loss should be the primary aim in type 2 diabetes, regardless of how it is achieved. In the rebuttals and questions, there was active debate about the benefits of low carb diets, and the audience became skeptical about the benefits of depleting glycogen. The debate was about promoting a high carbohydrate diet, but the consensus seemed to be that you need enough carbs to be physically active. In the final vote, a strong majority agreed that it’s time to stop promoting starchy carbohydrates and focus on (even modest) weight loss, even though there was no clear consensus on the merits of low carb diets.
As usual, Diabetes UK featured quite a large exhibition for the size of the conference. Here are our notable highlights:
- Janssen: The company (part of the J&J family) launched Invokana last week in the UK, so the exhibition was a coming out party. The team was also very happy with the glowing front-page coverage it received from a UK national newspaper the day after the launch. On the booth, they were aggressively emphasizing advantages on A1c, blood pressure and weight against Merck’s Januvia (sitagliptin) and also went in swinging against AZ’s Farxiga/Forxiga (dapagliflozin), based on canagliflozin’s head-to-head data against Januvia, its two doses, and its availability for patients with lower kidney function. Janssen were also showing official UK pricing for Invokana – at £1.31 (~$2.18) for 100mg (daily price) and £1.67 (~$2.77) for the 300mg dose. This compares to £1.31 (~$2.18) for Forxiga and £1.19 (~$1.98) for Januvia, although we understand actual prices paid can be different. Janssen was joined on the J&J booth by Animas, celebrating the pediatric approval for the Animas Vibe/Dexcom G4 Platinum.
- Takeda: The company was officially launching its DPP-4 inhibitor franchise Vipidia (alogliptin) and Vipdomet (alogliptin/metformin) at Diabetes UK for the UK market. They were proud of their proven two-year superiority to sulfonylurea (on top of metformin) and their excellent cardiovascular safety data in high-risk patients. They also noted that Vipidia was priced “to offer significant savings” – namely that it was roughly 20% cheaper than Merck’s Januvia or BI/Lilly’s Trajenta for a savings of around £86 (~$140) per year. Vipdomet was also priced the same as Vipidia, so the metformin is offered at no extra cost.
- Roche: The booth featured the Accu-Chek Mobile, an all-in-one meter that holds a cartridge of 50 test strips (that can be disposed of in the regular garbage), together with a lancing device and meter in a single device. Roche was suggesting that the meter was particularly of interest for motorists, who should be following the UK’s DVLA guidelines and testing before they start to drive. Roche was also displaying its Accu-Chek Insight next-gen pump + touchscreen meter remote, which the team is planning to launch in Europe over the next few months; it is a rather sleek-looking device.
- Lilly & Boehringer Ingelheim (BI): The alliance partners had a busy booth featuring the rapid-acting insulin analog Humalog (insulin lispro), the DPP-4 inhibitor Trajenta (linagliptin), and the partnership between Lilly and Disney that is coming to the UK. Lilly Diabetes was also displaying a wide range of educational materials for children at various ages.
- Bayer: The company was displaying its full range of Contour meters (Next Link, Next USB, Next, XT)
- Novo Nordisk: Fielding a relatively modest booth, Novo Nordisk was treating doctors to a virtual reality experience. The booth featured the GLP-1 agonist Victoza (liraglutide) and the ultra-long acting basal insulin Tresiba (insulin degludec), although we understand that Tresiba is very hard to get reimbursed in the UK at present (read our Novo Nordisk 4Q13 Report for more thoughts on Tresiba’s pricing).
- BD: The booth featured a big focus on lipohypertrophy, which leads to glycemic variability and unexplained hypoglycemia. The company is highlighting its 4 mm needle as a way to reduce the risk of lipohypertrophy, which we believe would be of major benefit to patients with this affliction.
- AstraZeneca: The company’s booth featured its broad portfolio of drugs for type 2 diabetes – the SGLT-2 inhibitor Forxiga (dapagliflozin), the DPP-4 inhibitor Ongylza/Komboglyze (saxagliptin), and the GLP-1 agonists Bydureon (exenatide once weekly) and Byetta (exenatide). Their materials were celebrating the fact that Forxiga had been recommended by NICE.
- Merck: The booth had an entertaining “Spotters Guide” to persuade doctors to spot potential patients that could see benefits with the DPP-4 inhibitor Januvia (sitagliptin). In particular, the patient profiles were those using sulfonylureas where hypoglycemia could be a problem (in the elderly, in a taxi driver).
- CellNovo: The company was riding the momentum of its recent UK insulin pump launch. Many in the US are waiting to see the next step; we have heard favorable things about the customer service in particular.
- Dexcom Distributors UK: Representatives told us that there were about 400 patients using standalone Dexcom in the UK, of which 90% were self-pay. The company didn’t have data on sensors used with the Animas Vibe, since they were distributed via Animas.
-- by John Close. Manu Venkat, and Kelly Close