Diabetes Victories Due to Improved Care Rather Than New Insulins
MANCHESTER, UK — The specific choice of insulin formulation, including whether to use a biosimilar, or "copycat," version of an established molecule, should be based on factors such as individual patient characteristics and cost, as there is little difference in performance between preparations, UK experts told the Diabetes UK Professional Conference 2017 here.
In a dedicated session on the efficacy, cost, and benefits of the newer injectables in diabetes management, David Levy, MD, of the London Diabetes Centre, United Kingdom, said that the impressive reductions in HbA1c levels and hypoglycemic episodes achieved in recent years have not been due to the availability of novel insulin molecules and preparations but to improved care and new technologies such as continuous glucose monitoring (CGM).
Indeed, performance differences between insulin formulations are largely negligible, Philip Newland-Jones, MSc, pharmacist for diabetes and endocrinology, University Hospital Southampton NHS Foundation Trust, United Kingdom, told attendees. He added that pharmacokinetic characteristics affecting administration and cost are more relevant factors in medication choice, with insulin merely a "tool" in the overall care package.
The session was chaired by Jeffrey W Stephens, MD, PhD, of the diabetes, endocrinology, and general internal medicine division at Morriston Hospital, Swansea, United Kingdom.
The trick, Dr Stephens told Medscape Medical News, is in "tailoring insulin choice to the patient in front of you and the traits they have. Taking into consideration whether they can manage two- or four-times-a-day insulin, whether there's a problem with…injecting, and so on, and whether they are at higher risk of hypoglycemia or if their eating habits are erratic.
"Clearly if there is a problem with hypoglycemia and erratic eating, then you're probably looking at an insulin analogue with a longer duration of action, which is obviously with a lower risk of hypoglycemia."
Noting also that safety with novel preparations and biosimilars is "always a hot topic," Dr Stephens said studies so far have been relatively short term and that data on longer-term safety are needed. But "observational studies and real-world studies haven't shown any adverse, untoward effects that we haven't expected."
Insulin Use: "We've Come a Long Way"
Dr Newland-Jones began his presentation with a look at the history of insulin use, saying, "We've come a long way in a number of years."
What has been seen more recently is "an explosion of trying to be 'harder, better, faster, stronger'; trying to be more physiological, basically. We're just about getting there, but we're still not matching where we need to be yet," he observed.
Discussing biosimilar insulins, he noted that, from a regulatory perspective, what is required is that they demonstrate high similarity in clinical trials to their originator molecules and have no clinically meaningful differences in immunogenicity or occurrence of adverse effects.
Trials of insulin glargine biosimilar (Abasaglar; Basaglar in the United States, Eli Lilly) comparing it with the originator product (Lantus, Sanofi-Aventis) have demonstrated this, he remarked.
"I think we can happily say, from the studies that Lilly has done, in terms of biosimilarity, that they have comparable clinical and safety profiles," he said, noting that the biosimilars come in around 15% cheaper because their development costs are less than for an originator molecule.
However, he emphasized that the responsibility for the choice of insulin product rests with the clinician in consultation with the patient, and the UK Medicines and Healthcare Products Regulatory Agency (MHRA) advises that dose adjustment may be needed with biosimilars.
Turning to concentrated insulins, Dr Newland-Jones said there has been "an explosion" of these products in the past few years, and he went on to list the various short-, long-, and mixed-acting concentrated insulins that are currently on the market.
The theoretical benefits of concentrated insulin are that they have altered pharmacodynamics and reduce the number of injections, the degree of pain on injection, and the overall insulin dose.
However, concerns about these higher-dose formulations include the possibility of prescribing errors, the potential for accidental overdosing, the fact that they require an appropriate device to administer them, and lack of education about them among healthcare professionals, diabetes patients, and their caregivers.
Insulin Is Just a Tool in the Whole Performance Package
Comparing specific concentrated insulins as examples, Dr Newland-Jones said studies on once-daily Toujeo U300 (insulin glargine, Sanofi-Aventis) indicated that it is noninferior to regular insulin glargine U100 (Lantus) in terms of HbA1c reduction, with no differences in hypoglycemia, at 6 months. One randomized controlled trial showed that body weight increased less with Toujeo than with Lantus, at a mean reduction in increase of 0.5 kg (P = .037).
Meanwhile insulin degludec (Tresiba, Novo Nordisk) — which is a longer-acting insulin with a half-life of around 25 hours — has shown some differences in performance in comparison with regular glargine U100, such as reduced rates of hypoglycemia and lower insulin doses, but these have not been consistent across studies, Dr Newland-Jones pointed out.
The main differences between formations such as Tresiba and Toujeo are in their pharmacokinetics and thus their dosing flexibility, he stressed.
And is some markets there may be cost considerations — for example, in the United Kingdom, a small difference in cost per 100 units results in a small, but not irrelevant, advantage for Toujeo since it is about £118 (~$140) per patient per year cheaper.
He went on to discuss Lilly's highly concentrated Humulin R U500, which he noted is not approved for use in the United Kingdom but is used there in certain "pockets." The few case series that have been published indicate that it achieves HbA1c reductions but at the expense of weight and hypoglycemic episodes. There are also concerns with Humulin R U500 in terms of the risk of prescribing error and overdosing, with prescribing usually retained by a specialist.
However, he explained that his center has achieved good results with Humulin R U500.
"Is this to do with the insulin?" he asked. "No. We spent a lot of time with these patients. They get one-to-one specialist multidisciplinary treatment, they get everything else as part of this package. It's not just the insulin. However, the insulin as a tool can be very effective, and I think that's the difference."
He added that, in his opinion, "The evidence for concentrated insulin is still lacking in the population; we might have perceived a benefit, but we haven't actually got that robust data from the companies.…There's ongoing responsibility for pharma …to make this as safe as possible."
Wrapping up, he said: "In summary, many new insulins and combination products give the opportunity for benefit, but they need to be part of the whole package. Medicines are really only tools, so you absolutely have to educate and support to utilize the full benefit."
Hypoglycemia: A Limiting Factor? Not Any Longer
Opening his presentation, Dr Levy said: "The major limiting factor of good glycemic control is hypoglycemia." This, he observed, "is a typical statement that you'll see all over."
However, data by Rosenbauer et al, published in 2012 (Diabetes Care. 2012;35:80-86), demonstrated that HbA1c levels in German and Austrian children with type 1 diabetes had "fallen by about 0.04% per year, amounting to an astonishing…achievement, which is a reduction in A1c…from a mean of 9.0% to 8.4%," yet at the same time rates of severe hypoglycemia had also fallen, by about 5% per year over the same period (from 1995 to 2009).
The question now is, has this relationship between glycemia and hypoglycemia changed over time?
Pointing to a more recent paper by Karges et al (Pediatr Diabetes. 2015;18:51-58), Dr Levy said that a "pseudo-DCCT curvilinear type relationship between 1995 and 2003" has been replaced by a completely flat line for the period 2004 to 2012.
He said: "There is no longer any relationship between hypoglycemia and achieved HbA1c in Germany and Austria in these young people. And let me just remind you that, at 6.5%, the rate is a quarter of what it was in the DCCT."
Crucially, these findings of a flat-line relationship between glycemia and hypoglycemia have also been replicated in the Type 1 Diabetes Exchange Registry in the United States and the Western Australia Childhood Diabetes Database, as described by Haynes et al (Pediatr Diabetes. 2016; DOI:10.1111/pedi.12477), and have been shown in patients using insulin pumps.
"Have insulin analogues contributed to this new epidemiological relationship?" Dr Levy wondered. "Of course, we can't do a definitive prospective study, but the authors of [Karges et al] thought that insulin was unimportant.
"They felt that noninsulin factors are now the major determinants of hypoglycemia risk and…very important and quite rightly, they include their own collaborative quality-improvement program as being a real-time indicator, back to centers, of how they're doing," helping to drive improvements across the whole system.
In support of this notion are data from trials comparing glargine (Lantus) with NPH insulin and insulin degludec (Tresiba) with glargine, which have shown that long-acting insulin analogues do not improve HbA1c or reduce overall hypoglycemia in type 1 diabetes and are not responsible for the large population reductions, he observed.
Technology, Care System Improvements More Important Than Type of Insulin
Dr Levy did, however, emphasize the urgent need to harmonize hypoglycemia definitions in clinical trials and for the inclusion of additional elements of care that have been shown to have a benefit, such as use of CGM.
He next discussed a 2016 Cochrane review of fast-acting insulin analogues vs human soluble insulin, which indicated that the former achieved no meaningful reductions in HbA1c or severe hypoglycemia and that there was not only a detectable performance bias but also a lack of robust studies on quality of life (Cochrane Database of Syst Rev. 2016;6:CD012161).
For the faster-acting insulin aspart (Novo Nordisk), a recent study does suggest that it has a greater glucose-lowering effect than regular insulin aspart after meals in patients with type 1 diabetes (Diabetes Technol Ther. 2017;19:25-33). But Dr Levy said that unpublished data indicate that there is no difference in type 2 diabetes.
Also in type 2 diabetes, a comparison between glargine and NPH revealed few differences, apart from a small reduction in nocturnal hypoglycemia of 0.4 episodes per patient per year.
Another study comparing Tresiba and Lantus also found a reduction in nocturnal hypoglycemic episodes, although Dr Levy pointed out that 85% of patients did not have such episodes and the reduction translated into one episode every 5 years vs one episode every 3 years.
"I think we'd be hard-pressed, even in our busy clinics, to discern any clinical difference," he stressed.
"Hypoglycemia is no longer a barrier to good glycemic control in type 1 patients; analogues have neither reduced overall hypoglycemia nor improved control and therefore haven't contributed to the extraordinarily impressive win–win seen in young people," he observed.
"Is there anything really new? Well, I have to say that there's not nearly as much as we all hoped."
In summary, he said: "Care system improvements are vastly more important than the insulins we use. Let's champion technology that really works — CGM, I might point out, would be a very good start."
The authors report no relevant financial relationships.
Diabetes UK Professional Conference 2017. March 8, 2017; Manchester, United Kingdom. Session