Held over three days in an inevitably wet Manchester, UK, and titled ‘The Future of Diabetes: Putting Evidence into Practice’, the 2013 Annual Professional Conference of Diabetes UK (the principal British patient, healthcare professional and research charity) provided a nicely balanced mix of basic and clinical science. The topics spanned such seemingly disparate fields as psychology, informatics and therapeutics. This diversity of subject matter was entirely apposite; such are the myriad facets of clinical diabetes. In this sense, the conference provided a timely reminder of the need to consider the entire individual; the days of clinical practice spent focusing myopically on tweaking the blood glucose level with the addition of another pill or a few units more of insulin have – thankfully – passed. Indeed, with the publication of recent major clinical trials, we have come to understand that lower blood glucose levels per se do not guarantee better long-term clinical outcomes. The latter has become a huge controversy that lies beyond the scope of the current article.
This report summarizes some of the sessions from the Diabetes UK meeting. It cannot hope to be comprehensive and so I offer selected highlights. I freely admit that these reflect my own professional biases, which no doubt will be readily evident. My aim has been to provide a report that is relevant to healthcare professionals who seek to enhance the health and well-being of those living with diabetes and related metabolic illness. By intent, the focus is on older individuals with diabetes, reflecting my academic interest in this age group.
Using my Editor-in-Chief’s prerogative I have granted myself the liberty of injecting one or two personal observations on the topics discussed, informed by three decades of personal professional involvement in the clinical care of people with diabetes.
Depression and diabetes
The enormous – but, it should be acknowledged, still rather underappreciated – psychological burden associated with diabetes was placed centre-stage in the opening plenary session of the conference. Although some endocrinologists might be more excited by updates about new glucose-lowering drugs (more of which later), less obvious – but very real – barriers to optimal care must be acknowledged and addressed. Only by doing so will people with diabetes derive the full benefit of any advances in pharmacotherapy. We must recognize the fact that diabetes is, par excellence, a long-term condition that requires an immense degree of self-management; diabetes healthcare providers can play only a supporting role.
The links between diabetes, depression and associated diseases are profound: diabetes doubles the risk of depression 1, whereas depression per se doubles the risk of developing type 2 diabetes. When the burden of the daily rigours of living with diabetes are coupled with the reality or threat of devastating tissue damage, it seems almost inevitable that depression should be the prime psychological ‘complication’ of the disorder. Yet, when I trained in diabetes – at arguably the largest diabetes clinic in the UK – depressive symptoms among our many patients were conspicuous by their absence. Was diabetes-associated depression less prevalent in the 1980s? An interesting possibility, but not easily tested. Or could it have been that the busy clinic teeming with patients and doctors (several to a single consulting room!) stifled the opportunity to explore the possibility? At a more fundamental level, with the notable exceptions of (a) patients with the sometimes extraordinarily unpleasant diabetic neuropathic symptoms and (b) the recurrent admissions of a small group of teenage girls in diabetic ketoacidosis, collective professional considerations of the psychological impact of diabetes were not as well developed as they are today; we live in somewhat more enlightened times. With regard to the wilful omission of insulin treatment by some of our younger patients, we remember (well, some of us perhaps) the convolutions of clinical scientists in the 1970s who were determined to prove the existence of an elusive ‘subcutaneous insulin resistance syndrome’. In reality, the explanation was more mundane, yet in some ways much more challenging, to the clinician and equally deleterious to the patients whom we tried to gently support as best we could.
Does the bidirectionality of the relationship between depression and diabetes contain clues to the prevention of and best treatment against diabetes? Professor Richard Holt of the University of Southampton, UK, quoted the 17th-century physician Thomas Willis, who observed that:
‘Diabetes is a consequence of sadness or prolonged sorrow’.
Although clinical science has since provided insights into potential psychoneuroendocrine mechanisms that connect the disorders, one wonders what led Willis to his prescient speculation.
Testosterone replacement in ageing men
This ‘Meet the Professor’ session featured Professor Fred Wu of the University of Manchester, UK. Professor Wu is an international authority on the impact of testosterone deficiency and long-term health outcomes in men. Latterly, this has become a highly charged and controversial area in therapeutics. TV viewers in the USA (where I currently reside) are subjected to an incessant bombardment of direct-to-patient adverts, extolling the benefits of testosterone replacement therapy – if you happen to be a lean, elegantly greying, middle-aged, happily married male model at least! The TV ad stereotype is well defined, which is rather more than can be said about the syndrome of age-associated male hypogonadism. Professor Wu presented data demonstrating the difficulties of defining what is an increasingly commonly recognized, but still rather indistinct, syndrome 2. A thorough clinical assessment must include not simply androgen levels but a broad symptom complex that includes perceived cognitive difficulties alongside adverse changes in body composition and impaired sexual function.
Decreased testosterone levels in middle age, in the absence of clearly identifiable causes, are surprisingly common. Epidemiological studies suggest that the risk of developing type 2 diabetes is significantly elevated in hypogonadal men. Testosterone replacement may bring symptomatic benefits, but the risks associated with long-term therapy, which include a theoretical increase in the incidence of prostate cancer, remain incompletely delineated. A clinical trial in Australia, now getting underway, may help to answer some important questions of risks versus benefits, notably whether testosterone replacement reduces the probability of type 2 diabetes.
Closely allied to this hypothesis is whether testosterone supplementation in men with unequivocal hypogonadism has a positive benefit on the future risk of cardiovascular events. The nature of the complex association between male hypogonadism and vascular disease was thoughtfully considered by Dr Martin Rutter of the University of Manchester at a scientific symposium organized by Cardiovascular Endocrinology3. Dr Rutter rigorously applied the causation criteria developed by Sir Austin Bradford Hill, the renowned 20th-century British medical statistician and epidemiologist. In brief, the current picture must be considered far from complete. Given the apparently increasing prevalence of age-related male hypogonadism, the need for well-designed clinical trials with meaningful clinical outcomes is clear.
Cancer, metformin and diabetes
Age and diabetes are recognized risk factors for certain types of cancer. Observational studies suggest that the well-established insulin-sensitizing drug metformin, which has been in use for more than half a century, may reduce cancer risk by approximately one-third. However, when Professor Rury Holman and colleagues at the University of Oxford, UK, carefully examined the data from published randomized controlled trials, they were unable to confirm any protective effect 4.
Of course, this kind of discrepancy – apparent benefits seen in observational epidemiological studies and subsequent failure to confirm such findings in randomized trials – has been seen before, famously when considering the health impact of postmenopausal oestrogen replacement therapy. This takes us back to Bradford Hill’s criteria of causality. With this in mind, Professor Holman described a new clinical trial, Glucose Lowering In Non-diabetic hyperglycaemia Trial (GLINT) – a randomized controlled trial to establish the effectiveness and cost-effectiveness of metformin in preventing cardiovascular events in people with nondiabetic hyperglycaemia at high risk for over 5 years (http://apps.who.int/trialsearch/trial.aspx?TrialID=ISRCTN34875079). The aim is to recruit 500 individuals over the age of 40 years who have high glucose levels that fall below the threshold for diagnosing diabetes. Participants will be allocated to receive metformin or placebo, and health outcomes, including cancer as a prespecified secondary outcome measure, will be followed up through medical records.
New drugs for diabetes
During the last 12 months, three new glucose-lowering drugs have been launched in the UK. Taken in the reverse order – and much in evidence in and around the conference – is lixisenatide.
This is the latest member of the class of glucagon-like peptide-1 receptor antagonists. In parentheses at this point, I should mention Dr Eugene Hughes, a primary-care physician from the Isle of Wight off the south coast of England who entertained the audience of healthcare professionals with thoughtful and witty observations on ‘The Language of Diabetes’. Dr Hughes prompted attendees to think about the words they use in everyday conversations with their patients (or ‘people with diabetes’ as many would prefer). For example, why do we seem to use the term ‘failure’ so often? We talk of ‘failure to reach glycaemic targets’ (‘whose targets’ anyway he asked?); we speak of ‘failure of oral therapy’, heralding the need for insulin. Certainly, as professionals we have a challenge on our hands when it comes to discussing new treatments with classes like ‘glucagon-like-1 receptor agonists’ and ‘sodium–glucose cotransporter-2 (SGLT-2) inhibitors’. Lixisenatide (Lyxumia) is a once-daily injectable agent for type 2 diabetes that joins the other members of the class, namely, exenatide (Byetta and Bydureon) and liraglutide (Victoza). These drugs have been well received by patients and endocrinologists, not least because they offer the prospect of weight loss in addition to improved glucose control (although not all people respond well). However, the class has also been under the shadow of safety concerns that were reignited during the conference with the announcement of a new review by the Food & Drug Administration (FDA). In the February 2013 issue of the Journal of the American Medical Association Internal Medicine, researchers at Johns Hopkins School of Medicine published a study revealing that people taking Byetta or Bydureon had double the rate for acute pancreatitis 5. Moreover, another drug in a different but related class – Januvia, which is an oral dipeptidyl peptidase-4 inhibitor – also appeared to share this increased risk. On 14 March 2013, the FDA confirmed that it is to investigate reports of possible increased risk for pancreatitis and precancerous findings of the pancreas from incretin mimetic drugs for type 2 diabetes. The European Medicines Agency (EMA) has since followed the FDA’s lead and has announced its own investigation.
Commenting on the pancreatitis paper in the British Medical Journal (BMJ 2013; 346:f1263), Edwin Gale, Emeritus Professor at the University of Bristol, UK, posed the question:
‘Should we be worried about this?’ His view: ‘Very much so. All forms of pancreatitis, clinical or subclinical, predispose to carcinoma of the pancreas.’
Prof. Gale also noted that the number of reports on pancreatitis received by the FDA had reached ‘astronomical proportions’. This has been a highly charged issue for several years, and views are somewhat polarized between key opinion leaders who believe that there is a genuine risk and others who are clearly much less convinced.
Insulin degludec is an ultra long-acting insulin analogue launched in the UK in January 2013 by NovoNordisk under the brand name Tresiba. It is available in two strengths (standard 100 U/ml and double strength 200 U/ml) in preloaded pens. The duration of glucose-lowering action is in excess of 42 h at clinically relevant doses. The unique properties of this analogue permit greater flexibility in the timing of daily dosing. A daily injection can be given as early as 8 h or as late as 40 h after the previous dose. Insulin degludec has been shown to reduce the incidence of nocturnal hypoglycaemia compared with the market leader, Lantus (insulin glargine). However, although the EMA approved insulin degludec back in 2012, the FDA subsequently stunned NovoNordisk earlier this year by demanding that the company perform a pre-approval cardiovascular safety outcome study. So here we have the uncomfortable situation wherein one regulator (EMA) approves the drug, whereas the FDA – assessing the same data – expresses major misgivings about a potential increased risk for cardiovascular events. This decision is expected to delay the approval of insulin degludec in the USA by several years. A view expressed by senior investigators in the UK is that US regulators are being particularly – perhaps read ‘unduly’ between the lines – cautious about granting approval to new diabetes drugs because of the recent safety debacle surrounding rosiglitazone 6. In brief, far from preventing heart attacks as had been hoped (and hyped) after a decade of clinical use, rosiglitazone was found to be associated with an increased risk for this very complication – the leading cause of premature death among people with type 2 diabetes. Although it can be argued that the reanalysis of the rosiglitazone safety data did not provide unequivocal evidence of harm, the fact remains that the claim could not be disproved. In the wake of the rosiglitazone affair, both the FDA and EMA issued guidance about the need to establish minimum levels of cardiovascular safety for new glucose-lowering drugs. Taking an even longer historical perspective, there are precedents for diverging views on cardiovascular safety of diabetes medications. The widely used sulfonylureas have long carried a warning in the USA about a potential increase in risk for major cardiovascular events such as heart attack. This stemmed from a US clinical trial completed in the early 1970s that European clinicians regarded as deeply flawed, to the point of ignoring the concerns raised. Anyone interested – and that is all of us, by the way! – in reading more about the sometimes tense relationship between drug companies and regulators (and doctors) should avail themselves of UK academic Dr Ben Goldacre’s excellent recent book ‘Bad Pharma’ 7. There are very recent hints that things are starting to change; they need to, for the sake of the credibility of medical science 8.
This orally active drug from AstraZeneca and Bristol-Myers Squibb was launched in November 2012 in the UK. It is the first of an entirely new class – the aforementioned SGLT-2 inhibitors. Canagliflozin (Invokana from Johnson & Johnson) was approved by the FDA on 29 March 2013, the agency having rejected dapagliflozin (see below). Several more agents in this class are under development. Endocrinologists have had to adjust their mindsets with the arrival of the SGLT-2 inhibitors. After all, traditional teaching had always been to seek to reduce the amount of glucose in the urine of patients with diabetes, the main objective being to relieve osmotic symptoms. Freedom from glucosuria was the main evidence of successful therapy in the not-so-distant days before haemoglobin A1c levels and self-testing of capillary glucose levels became widely accepted as indicators of glucose control. Therefore, to now start using drugs that lower blood glucose by increasing the amount of glucose excreted in the urine requires that the traditional notions of glucosuria be, we might say, ‘revised’ (with some older physicians probably recoiling in disbelief). Certainly professional opinions are divided, that much is clear. We find ourselves in another uncomfortable situation wherein the success or otherwise of new diabetes drugs nowadays depends in part on negative perceptions of the safety and tolerability of the alternatives. There is renewed debate about the cost-effectiveness of some of the newer classes of glucose-lowering drugs, especially given their generally modest impact on glycaemia. Although an expanding range of options chimes with the current approach of ‘individualizing’ pharmacotherapy for patients with diabetes, concurrently we are witnessing a dramatic increase in the range of potential adverse effects, only some of which are clinically apparent at this time.
The SGLT-2 inhibitors can theoretically be combined with a wide range of other glucose-lowering therapies. Additional attractions include an intrinsically low risk for hypoglycaemia and the prospect of weight reduction resulting from urinary calorie losses. However, known by the brand name Forxiga, this first-in-class agent presents yet another example of wildly differing perceptions between European and US regulatory authorities on what constitutes a risk to patients. In late-phase clinical trials, dapagliflozin was found to be associated with an excess of cancers compared with placebo. Nine breast cancer cases were observed in the dapagliflozin group and one case in the control group; moreover, there were nine cases of bladder cancer in the dapagliflozin group and one case in the control group. This ‘imbalance’ (as it was termed) appears to have no firm biological plausibility. In other words, the fact that more of these cancers were diagnosed in patients randomized to receive the drug could have been an unfortunate play of chance, a random event not in reality related to the drug. This takes us back to the discussions above about how true causality can be proved or disproved in such situations – which in reality could perhaps be summed up by the words ‘rarely definitively’. Given the fact that diabetes is now increasingly being regarded as a risk factor in its own right for certain types of cancer (see above), the dapagliflozin trial results are clearly a cause for concern and come closely after the revelation about a small increase in risk of bladder cancer with long-term use of pioglitazone. Doctors and healthcare professionals alike are left to ponder on how to reconcile the different approaches taken by the EMA and FDA. Clearly, this is a less than satisfactory state of affairs that many – healthcare professionals and patients – might be forgiven for feeling is in urgent need of resolution.
Risk–benefit analyses by regulatory authorities often have a subjective component. Although, as already mentioned, the FDA recently approved another drug in this class – canagliflozin – the recommendation came amid concern about an increase in cholesterol levels, cardiovascular events (especially stroke), renal function and bone safety. The SGLT-2 inhibitors may prove to be of value in older people with diabetes. However, warnings have also been sounded on the need for caution on the risk of genitourinary infections, particularly in ‘elderly, frail patients and in patients with impaired kidney function and/or high cardiovascular/cancer risk who represent an increasing fraction of the population with diabetes mellitus’ 9. In draft guidance issued in February 2013, the National Institute for Health and Clinical Excellence (NICE) said that it was ‘minded not to recommend’ Forxiga as an add-on therapy to other agents, citing ‘significant uncertainty about the validity of the results’ of trial data submitted by the manufacturer. A final decision is expected in June 2013.
Conflicts of interest
The views expressed are those of Dr Krentz and do not necessarily reflect those of other members of the editorial board, Lippincott Williams & Wilkins nor any other commercial or academic institution. Dr Krentz has contracted as either a consultant, advisor, speaker and/or has received research funding from pharmaceutical companies including Abbott, Astra Zeneca, Bayer, Boehringer-Ingelheim, Bristol–Myers Squibb, Eli Lilly, GlaxoSmithKline, Halozyme, Johnson & Johnson, Merck, Novartis, Novo Nordisk, Pfizer, Roche, Sanofi-Aventis, Servier and Takeda.