The 75th Anniversary of the American Diabetes Association was a noteworthy conference that had plenty to interest the 18 000 or so attendees of the scientific sessions in Boston.
Effect of glucose-lowering drugs on cardiovascular outcomes
The headline papers from a metabolic-cardiovascular perspective were TECOS (Trial to Evaluate Cardiovascular Outcomes after Treatment with Sitagliptin) and ELIXA (Evaluation of Lixisenatide in Acute Coronary Syndrome). Both have been extensively reported on and only a brief recap is offered here. In TECOS, among patients with type 2 diabetes and cardiovascular disease, the dipeptidyl peptidase-4 (DPP-4) inhibitor sitaglipitin was not associated with an increased risk of cardiovascular events. This result argues against the heart failure signal previously observed with saxagliptin in SAVOR-TIMI 53 as being a class effect of DPP-4 inhibitors. As TECOS was designed as a noninferiority study, but had a sufficient statistical power for a superiority analysis, the result also dampens any residual hopes of specific cardiovascular advantages with this class of glucose-lowering agents. In ELIXA, the recently approved injectable GLP-1 agonist lixisenatide did not show any benefit on cardiovascular outcomes in 6000 high-risk patients with type 2 diabetes. However, expert commentators seemed reassured that lixisenatide represents a safe therapeutic option. It should also be noted that, yet another analysis of the long-debated cardiovascular safety of a much older class of glucose-lowering drugs – sulfonylureas – found no evidence of harm; this finding stands in contrast to other published analyses.
The selection of sessions that follows to some extent reflects my personal interest in the various topics while retaining a broad perspective that is in alignment with the aims of Cardiovascular Endocrinology. I offer this report in the hope that it may be of interest to readers of the journal who perhaps may not have explored the evidence, such as it is, on the novel supportive care that is increasingly available for patients. If used effectively in conjunction with pharmacotherapy – which often is not the case in clinical practice – this approach has the potential (yet to be proven) to improve cardiometabolic outcomes.
New and emerging pharmacotherapies for diabetes
Biosimilar formulations of insulin glargine and insulin lispro reportedly met equivalence targets in euglycaemic clamp studies (see below). Early data from once-weekly insulins, which have potential to be combined with weekly GLP-1 agonists, were also presented. Predictably, much new or confirmatory data were presented on GLP-1 receptor agonists, DPP-4 inhibitors and sodium glucose cotransporter-2 (SGLT-2) inhibitors. Reports of diabetic ketoacidosis have recently cast something of a cloud over the latter class. This concern has prompted reviews by the Food and Drug Administration (FDA) (http://www.fda.gov/Drugs/DrugSafety/ucm446845.htm) and EMA (http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/referrals/SGLT2_inhibitors/human_referral_prac_000052.jsp&mid=WC0b01ac05805c516f). Safety was also a consideration for the novel long-acting insulin peglispro with concerns focused on increased liver fat, triglycerides and elevations of hepatic transaminases.
Making sense of technology
Kelly Close (http://www.closeconcerns.com, http://www.diatribe.org) started her lecture on new technology by reminding the audience that a given haemoglobin A1c (HbA1c) of, for example, 7.0% (53 mmol/mol) could reflect very different daily blood glucose profiles in individual patients. One patient might have ∼80% of blood glucose results within target range with little variability; another patient might have an identical HbA1c level that conceals much greater variability with frequent incursions into the hyperglycaemic and hypoglycaemic ranges. Ms Close, speaking as a person with type 1 diabetes, also stated her opinion that aspects of clinical trials in diabetes needed to be rethought, notably to enable participation of some of the many patients with psychological problems that create major barriers to effective and sustained self-care and who tend to be excluded from efficacy trials.
Novel technologies for behaviour change
The concept of healthcare versus self-care was echoed and expanded in other sessions. Diabetes is a complex and largely self-managed long-term disorder for which the majority of patients do not receive anything approaching what could be called adequate education. Although novel technologies may increasingly offer a partial solution to this ongoing problem, only 50% or fewer patients currently engage with such alternative approaches. The need for patient-centred technology that is easy to use and provides what the individual patient needs at the time (rather than what family or healthcare professionals may propose from their perspectives) was a recurring theme.
However, the statement by another speaker (Margaret Powers, PhD, of the American Diabetes Association, Alexandria, Vermont, USA) that glucose data per se drive improvements in metabolic control could perhaps be challenged. Appropriate and meaningful data are certainly required as a basis for informed decisions on nutrition, exercise and diabetes therapy, but data that are collected but not acted on are not necessarily of value; many suboptimally designed studies of self-monitoring of blood glucose in patients with type 2 diabetes attest to this.
Dr Chandra Osborn, PhD (Vanderbildt University, Tennesse, USA), detailed the difficulties of developing supportive technology in an academic setting. Problems of bureaucracy and limited resources contribute towards slow progress that risks being rendered rapidly obsolete by evolution in hardware and software.
The cost of insulin
Dr Irl Hirsch, MD (University of Washington, Seattle, Washington, USA), noted that the mechanisms for setting the cost of insulin are not transparent. The global insulin market is estimated to be $20 billion; Lantus leads the field in sales of insulin analogues. Dr Hirsch observed that although the price of insulin differs markedly between countries, it was hard not to believe that the manufacturers were making a loss when their products were sold at lower prices. However, he acknowledged that hard data are difficult to find.
Lutz Heinemann, PhD (Science and Co., Germany), expanded on the issue of varying insulin prices and pointed out a 5000-fold difference for a vial of insulin in Iran (cheapest) and Austria (most expensive). Professor Heinemann expressed his opinion that, in contrast to the general experience of generic drugs, the introduction of biosimilar insulins would be unlikely to markedly reduce the cost of insulin in developed countries. His estimate was a 20–30% price reduction on the basis of the history of other biosimilar products.
Diabetes devices and technologies
Continuing the theme of an earlier session Adam Brown, Head of Diabetes Technology at Close Concerns (see above), presented his views on what makes good diabetes technology from his vantage point as an individual with type 1 diabetes. Using the iPhone as his model, he identified several characteristics (makes life easier, significant improvement over available technology, ‘it just works’ i.e. no need for instructions, etc.) that may be used to define success. He chronicled the development of some relevant medical technology hits (OmniPod) and misses (GlucoWatch) as examples. In Mr Brown’s view, creating devices is ‘so hard’, with the focus too often being on navigating the Food and Drug Administration approval process and trying to make a product fit the market instead of starting with an analysis of market need, that is putting the patient first. Most importantly, he observed, diabetes technology should never increase the already considerable burden of living with diabetes.
Another speaker in this session noted that diabetes education has thankfully evolved to a more enlightened attitude that has moved away from the paternalistic ‘You should do this…’ approach that characterized diabetes care in the past. This change in attitudes by medical educators has spurred the development of numerous smartphone applications for diabetes. However, most such applications do not ‘stick’ as users lose interest and abandon the technology. During Q&A, a US-based clinician cautioned against placing too much reliance on new technology. In his opinion, the medical profession should strive to make attendance at specialist diabetes clinics a more rewarding – even attractive – experience. Although somewhat tongue-in-cheek, the point was well made: how many of our patients look forward to a visit to the endocrinologist’s office or a hospital clinic?
John M Jakicic, PhD (University of Pittsburgh, Pennsylvania, USA), presented an excellent overview of the importance and assessment of exercise in the prevention and treatment of type 2 diabetes, with a focus on the role of activity trackers. An articulate, convincing and witty advocate for exercise, Dr Jakicic emphasized that glucose-lowering pharmacotherapy often leads to suboptimal improvements in glycaemic control because activity levels of the individual are too low. A dose–response relationship exists between fitness and HbA1c levels. Intriguing unpublished data from his group showed that 10 min of exercise after a meal lowers postprandial glucose increase, thereby flattening the insulin secretory response. Thus, fitness may result in metabolic benefits potentially independent of any effects on obesity. This has to be an area worthy of further exploration. In Dr Jakicic’s view, wearable measurement devices are of value as facilitators of behaviour change, noting in passing that accelerometers are in fact ‘very old’ technology. His research group has validated accelerometers against indirect calorimetry as a measure of energy expenditure. He noted that among the commercially available devices, the best known (and most heavily promoted) are not necessarily the most accurate. Moreover, devices do not have to be ‘high tech’ to make a difference: the use of $10 pedometers can encourage increases in daily step counts by ∼2000. Dr Jakicic looked forward to the point where pattern recognition analysis of data collected from wearable devices might be used to predict health outcomes, for example impending hypoglycaemia/hyperglycaemia, for an individual. This, he noted, would mark the arrival of ‘smart’ technology. He also stressed the importance of not taking the counsellor (i.e. healthcare professional) out of the loop when using wearable technology to facilitate behaviour changes and presented convincing data from controlled clinical studies to support this contention. Another note of caution was sounded that exercise – which brings clinical benefits of relevance to cardiometabolic disease, for example on depression, cardiovascular risk profiles, and which is safer than most FDA-approved drugs – should not be overmedicalized.
A study presented in another session reinforced Dr Jakicic’s position on the long-term value of physical fitness. Levels of fitness at age 20 and increasing levels of fitness in middle age appear to protect against the development of type 2 diabetes. However, the latter finding was attenuated by adiposity, a finding that could be interpreted as detracting somewhat from Dr Jakicic’s comments.
Mice are from Mars, humans are from Venus – sex differences in the pathophysiology of diabetes
This session won my award for the least comprehensible title at the conference; perhaps a typographical error crept into the programme unnoticed. In fact, the session provided a good overview of the influence of sex steroid hormones on aspects of islet function, insulin action and risk of diabetes. Included were detailed comparative data on older and younger men and women from Dr Rita Basu (Mayo Clinic, Rochester, Minnesota, USA) showing that during hyperinsulinaemic euglycaemic clamps, significantly greater arterial concentrations of insulin were found in men. This was interpreted as reflecting greater peripheral insulin clearance in women. This observation has implications for the interpretation of glucose clamp studies as exogenous insulin infusion rates calculated on the basis of total body weight versus fat-free mass may result in sex-specific differences in achieved hyperinsulinaemia and hence in the apparent effect on glucose metabolism.
Endothelial function in metabolic regulation
In this state-of-the-art session, the well-described interactions between the microvasculature, microvasculature and insulin action were extensively reviewed. Dr Cathryn Kolka (Cedars Sinai Hospital, Los Angeles, California, USA) presented original research data that used a dog model of overfeeding. A given level of adiposity induced by feeding with animal fats induced insulin resistance and hyperinsulinaemia. By contrast, equivalent weight gain achieved using ω-3 fatty acids was associated with preserved whole-body insulin sensitivity including normal levels of plasma and interstitial (lymph) insulin.
What is the best method to quantify insulin sensitivity?
This question has been posed for decades and so was revisiting well-trodden ground. In this sense, the session might have been considered of somewhat historical interest and perhaps not so relevant to today’s clinical investigators. This was not the case. The symposium was presented by some of the best-known pioneers in the field. Perhaps these big names were part of the draw and the session was surprisingly well attended. The other reason for the large and attentive audience presumably reflects continuing interest in this topic – and continuing uncertainty about the enigma of insulin resistance and how it is best quantified. If I have a criticism it concerns the repeated use of the word ‘physiological’ in the context of what are all very pharmacologic approaches to measuring insulin’s effects on glucose metabolism.
Dr Robert Rizza (Mayo Clinic, Rochester, Minnesota, USA) reviewed the hyperinsulinaemic euglycaemic clamp, which is generally considered the reference method for quantifying insulin action in glucose metabolism. In his characteristically modest way, he offered important practical insights into glucose clamp methodology that he freely admitted had been discovered by ‘making many mistakes over many years’. Following publication of the classic paper by Dr Ralph DeFronzo, investigators who included Dr Rizza and Dr Jerry Olefsky enthusiastically applied and developed the technique. The early dose–response studies of these investigators have stood the test of time. However, the issue of precision, i.e. repeatability of clamp results, was mentioned only in passing. Dr Rizza explained that there were ‘few data’ in the literature on intraindividual variability of clamp measures. He offered sound practical advice to clinical investigators engaged in glucose clamp studies:
- ‘D50 is not D50’ – Action: measure the glucose concentration in the 50% dextrose infusate – and ‘do it in quadruplicate’.
- Check tracers for purity (recalling the days when radioactive-tritiated glucose was used to determine glucose turnover).
- Apply caution in using the Steele nonsteady–state equation if there are rapid changes in specific activity.
- M/I ratios (the glucose disposal rate divided by the prevailing serum insulin concentration) should not be used in comparative studies if plasma insulin concentrations vary appreciably between groups.
- It is necessary to clamp at identical blood glucose targets for between-group and within-group comparisons.
- The liver and adipocytes are exquisitely sensitive to small increments in insulin and are suppressed at the levels of the hormone typically attained in hyperinsulinaemic euglycaemic clamps.
- If attempting to calculate gluconeogenesis, take care and use appropriate tracers.
In Dr Rizza’s opinion, the insulin sensitivity glucose clamp should be best considered a tool to test hypotheses creating a basis for more detailed mechanistic studies, for example multiple oral and intravenous glucose tracers as developed in collaboration with his senior research colleagues at the Mayo Clinic.
Dr Gerald Reaven (Stanford University, Stanford, California, USA) reviewed the history of the study of insulin action in human disease spending time on the seminal studies of Sir Harold Himsworth in the 1930s. Even before insulin could be measured, these led Dr Himsworth to the realization that diabetes could readily be divided into insulin-sensitive (now known as type 1) and insulin-insensitive (type 2) subtypes. Dr Reaven rued the fact that he was quite unaware of Himsworth’s experiments throughout his training, even though they had been published in the Lancet, no less. As a junior researcher, and in the knowledge of insulin measurements made possible by Yalow and Berson’s development of the radioimmunoassay, Dr Reaven developed the insulin suppression test on the basis of the intravenous infusion of epinephrine, propranolol, insulin and glucose. Perhaps remarkably, the insulin suppression test (nowadays modified to use octreotide in place of epinephrine to suppress endogenous insulin secretion and the β-blocker to counter the adverse vascular effects of the epinephrine) compares very favourably with clamp-derived measures of insulin action. Dr Reaven presented robust data in the form of Bland–Altman plots in support of this contention. In Q&A, Dr Reaven was highly dismissive of the homoeostasis model assessment (HOMA)-IR index, which is calculated from fasting insulin and glucose concentrations. According to Dr Reaven, in nondiabetic individuals, the index is not superior to measurement of insulin and in the presence of hyperglycaemia HOMA-IR becomes ‘nonsensical’.
Dr Richard Bergman (Cedars Sinai Hospital, Los Angeles, California, USA) developed the so-called ‘minimal model’, which describes how insulin acts in the body and has become a widely used method in clinical metabolic research. Dr Bergman pointed to the mathematical model of the relationships between insulin secretion, insulin action and other metabolic factors that can be derived from data generated using the frequently sampled intravenous glucose tolerance test (FSIVGT). Drawing on his engineering background, he created a method that provides a comprehensive and integrated picture of metabolic function, albeit with certain caveats. An aspect of glucose metabolism that the minimal model offers is a measure of ‘glucose effectiveness’, that is, the ability of glucose per se to influence its own clearance. In this way, Dr Bergman proposed that glucose could be considered to act as a ‘quasi-hormone’. The success of the minimal model did not come overnight and, in fact, Dr Bergman’s key 1985 paper on the technique has only become widely cited during the last few years; the minimal model has certainly proved its durability. Dr Bergman spent some time reviewing the topic of insulin clearance (which occurs primarily in the liver) and presented new data using a new mathematical model that suggests that liver and peripheral insulin clearance are independently regulated. He also provided data proposing that liver hexokinase activity could be inferred by simultaneous measurement of glucose and lactate levels during the FSIVGT. Dr Bergman expressed incredulity at the continued use of a widely used (and misused) test (HOMA-IR) that he evidently considers as potentially misleading. Indeed, such was his scorn for HOMA-IR that he freely described himself as a ‘HOMAphobic’. Dr Bergman also had negative comments on some of the indexes of insulin action that have been applied to oral glucose tolerance test data. The Stumvoll index, for example, includes BMI, which Dr Bergman considered undesirable in a measure of insulin sensitivity.
Fifty years of diabetes research
Some 25 years ago, when I was engaged in my first period of clinical diabetes research, I had the pleasure of spending time as a visiting scholar with Dr Daniel Porte Jr (University of Washington, Seattle, Washington, USA) and some of his eminent colleagues. Dr Porte, who is currently professor of medicine at the University of California San Diego, presented a personal perspective on half a decade of research in diabetes.
- Epinephrine suppresses endogenous insulin secretion, a seminal study of Dr Porte’s, but in fact a somewhat unintentional finding removed from the hypothesis that was tested.
- Pathogenesis of type 2 diabetes – the relative contributions of insulin resistance versus impaired β-cell function.
- Insulins – the introduction of products with improved purity, monocomponent insulins, human insulin, insulin analogues.
- Physiological role of the incretin system (see below).
- DCCT (Diabetes Control and Complications Trial) – benefits and risks of intensive treatment of type 1 diabetes.
- UKDPS (United Kingdom Prospective Diabetes Study) – Ditto for type 2 diabetes.
- Identification of MODY (Maturity Onset Diabetes of the Young) genes.
- Metformin – approval in the USA only in 1995 following decades of use in Europe and elsewhere.
- Thiazolidinediones – which arrived ∼20 years after clamp studies showing the importance of insulin resistance in the pathogenesis of type 2 diabetes.
- Genome-wide association studies and diabetes risk.
- Brain regulation of weight and metabolism; the endocrine system and CNS are functionally connected.
- Incretin mimetics, SGLT-2 inhibitors, etc. Again, the development of effective drugs followed improved understanding of the relevant pathophysiology and the identification of therapeutic targets.
Dr Porte observed that most hypotheses in science turn out to be wrong; in his opinion, something of a suspension of belief is necessary. Moreover, fundamental research rarely leads directly to clinically relevant advances. Clinical research – especially focusing on physiology – is particularly difficult. He emphasized the critical role of career support for so many clinical scientists from the American Diabetes Association.
Some miscellaneous personal thoughts and observations
It is high time that presenters of state-of-the-art lectures were asked to provide a brief outline of their talks.
There was way too much use of the redundant and pointless phrase. ‘That’s a great question!’ prefacing responses from presenters during the Q&A session. This irritating time-wasting device has reached epidemic proportions.
Although credible efforts were made to encourage attendees to avail themselves of the opportunities for aerobic exercise (Photo 1), my observations led me to conclude that more than 90% declined to have anything to do with the staircases. Instead, the escalators were clogged by riders resolute in their refusal to place one foot in front of the other.
Conflicts of interest
There are no conflicts of interest.