Targeting Hypoglycemia and Glycemic Variability in T2D? More From DEVOTE

September 15, 2017

LISBON, PORTUGAL — A second wave of results from the DEVOTE trial — which pitted the ultralong-acting, once-daily basal insulin degludec (Tresiba, Novo Nordisk) against the commonly used insulin glargine in patients with type 2 diabetes at high cardiovascular (CV) risk — provide insight on the impact of glycemic variability upon severe hypoglycemia and CV outcomes and the temporal relationships between severe hypoglycemia, CV outcomes, and deaths in the study.

"These are new and prespecified secondary analyses," senior investigator of the overall DEVOTE trial, John B Buse, MD, PhD, of University of North Carolina School of Medicine, Chapel Hill, told the EASD meeting here today.

Full results of the Trial Comparing Cardiovascular Safety of Insulin Degludec versus Insulin Glargine in Patients with Type 2 Diabetes at High Risk of Cardiovascular Events (DEVOTE) were first reported at the American Diabetes Association (ADA) 2017 Scientific Sessions in June, showing showed that insulin degludec was equivalent in terms of CV safety with glargine but that the patients using the ultralong-acting insulin had a 40% relative reduction in severe hypoglycemia and a 53% relative reduction in nocturnal severe hypoglycemia compared with those using glargine.

First up to present the secondary findings was Bernie Zinman, MD, of Mount Sinai Hospital, Toronto, Ontario, who reported on the DEVOTE-2 analysis, which showed that day-to-day fasting glycemic variability was associated with increased risks of severe hypoglycemia and all-cause mortality in this patient population — but not with the outcome of major cardiovascular adverse events (MACE) once adjustment had been made for baseline characteristics and the most recent HbA1c. The DEVOTE-2 findings were published simultaneously in Diabetologia.

Next, Thomas R Pieber, MD, of the Medical University of Graz, Austria, reported an association between severe hypoglycemia and all-cause death in DEVOTE-3, also simultaneously published in Diabetologia. There was no significant association between severe hypoglycemia and MACE, however, although there was a significantly higher risk of CV death following a severe hypoglycemic event. The data, said Dr Pieber, "support a temporal relationship between severe hypoglycemia and all-cause mortality" and indicate that severe hypoglycemia is associated with subsequent death.

Invited to give independent comment was Martin Rutter, MD, FRCP, senior lecturer in cardiometabolic medicine at Manchester Royal Infirmary, United Kingdom, who has also penned an editorial accompanying DEVOTE-2 and DEVOTE-3, again simultaneously published in Diabetologia.

"While DEVOTE-2 and DEVOTE-3 raise awareness of the mortality risks associated with glycemic variability and severe hypoglycemia in high-risk insulin-treated patients with diabetes, they cannot clarify causal relationships," he writes.

"Preventing severe hypoglycemia in those with type 2 diabetes should already be a priority in clinical practice."

But findings from future clinical trials "are needed to guide physicians on whether it is beneficial to target glucose variability and risk for severe hypoglycemia to reduce the risks for CVD events" and death in these individuals.

Targeting Glycemic Variability: Not Yet Ready for Prime Time?

In presenting the DEVOTE-2 findings on glycemic variability, Dr Zinman explained that HbA1c "is not the only index" of glycemic control. "These data support a clinical benefit of a basal insulin that has low day-to-day variability and therefore provides consistent fasting glycemia," he said.

Dr Rutter said these new results "are important and timely," and he applauded DEVOTE-2 for being the largest study in high-risk insulin-treated type 2 diabetes patients to relate glucose variability to MACE, mortality, and risk for severe hypoglycemia, with "high-quality data and adjudicated events," but said there were nevertheless some limitations.

"These were observational data, and there were differences in baseline risk factors across glucose-variability groups," he noted. Therefore "high glucose variability might simply be a marker of other unmeasured conditions or behaviors leading to severe hypoglycemia and higher mortality." One example of such an unmeasured confounder could be autonomic neuropathy, he said.

There was also "limited statistical power" for some of the analyses, Dr Rutter added, and the glucose variability was obtained from only three fasting blood glucose results per month.

"Continuous glucose monitoring [CGM] would have captured more glycemic variability," he observed.

"If I'm being honest, it would seem premature to target glycemic variability to reduce severe hypoglycemia," Dr Rutter asserted.

He added that there are several ongoing trials in type 2 diabetes that have a reduction in glucose variability as the primary end point, but unfortunately "no studies target glucose variability to reduce cardiovascular events."

Reasons for this might be the practical issues of performing such trials, and as such there has been a lack of interest from pharmaceutical companies, which has probably not been helped by disappointing results from studies targeting postprandial hyperglycemia, including HEART2D and the Acarbose Cardiovascular Evaluation (ACE) study, reported here at the EASD meeting just 2 days ago, he said.

In conclusion, these new "strong observational data" show that fasting glucose variability is related to severe hypoglycemia and mortality and "help build a case for trials to determine whether targeting glucose variability reduces severe hypoglycemia, CVD risk, and mortality."

DEVOTE-3: We Don't Know if Severe Hypoglycemia Is Causing Death

DEVOTE is not the first CV outcomes trials in type 2 diabetes to show an association between severe hypoglycemia and increased rates of MACE and all-cause mortality — this has been previously been reported in ACCORD, ADVANCE, VADT, LEADER, ORIGIN, and EXAMINE.

Dr Rutter said the DEVOTE-3 results assessed temporal relationships of hypoglycemia to outcomes and again represent "high-quality data on adjudicated events in a large number of high-risk patients."

But, similarly, there are shortcomings. There was no assessment of the contribution of nonsevere hypoglycemia episodes nor of participants with multiple episodes of severe hypoglycemia.

The data are observational, and differences in baseline characteristics between those individuals with and without hypoglycemia could explain some of the relationships demonstrated: "Is the relationship causal or confounded by factors such as frailty? Is severe hypoglycemia acting as a marker?

"We do not know if those who died after severe hypoglycemia died because of severe hypoglycemia," he emphasized.

And there is no explanation from DEVOTE "of how physicians managed severe hypoglycemia to prevent further events. Was it in line with clinical practice?"

Dr Rutter also believes that it is "implausible that a single episode of severe hypoglycemia can have an effect on mortality 1 year later."

DEVOTE-3 "raises awareness of the mortality risks associated with severe hypoglycemia" and indicates that detection and prevention of hypoglycemia in type 2 diabetes is "not given the attention it deserves," he emphasized.

Overall, he indicated, the DEVOTE-3 data indicate that "when appropriate, there should be greater use of therapies for type 2 diabetes that lower the risk of hypoglycemia — including [sodium–glucose cotransporter 2] SGLT2 inhibitors, [dipeptidyl peptidase-4] DPP-4 inhibitors, [glucagonlike peptide 1] GLP-1 agonists, and fast-acting and ultralong-acting insulin analogues, including insulin degludec."

Finally, there are research implications, he noted — larger cohort studies with CGM and ECG monitoring "could provide important insights," and trials targeting risk for severe hypoglycemia "could quantify the benefits of intervention on hypoglycemia risk to reduce CVD and mortality risks."

Let's Test the Apps for Hypoglycemia and Let's Use Them

Dr Buse also presented what he emphasized was "a post hoc analysis" on a risk score for severe hypoglycemia developed using information on "well-known" risk factors from DEVOTE that are recognized as influencing the risk of having a hypoglycemic event. "We wanted to know if we could predict those in DEVOTE who might develop hypoglycemia further down the line."

It turned out that the long list of risk factors in DEVOTE had the same predictability for severe hypoglycemia as a shorter list of "traditional" risk factors — namely the insulin-treatment regimen (any, basal only, or basal and bolus), duration of diabetes, sex, age, and HbA1c at baseline — so this shorter list was used to develop a risk score, because these are "more easily translated into the clinic," said Dr Buse.

From this risk engine, they developed an app, which predicts "risk of hypoglycemia but also of MACE and all-cause mortality," he explained, adding that in those with the highest risk score for severe hypoglycemia in DEVOTE — when taking insulin degludec — there was a reduced risk of severe hypoglycemia and MACE.

Dr Rutter said the app is "a great idea! Let's test the apps and let's use them."

The authors have made slides from today's DEVOTE presentations as well as a QR code for the hypoglycemia risk score app available online at

This trial and secondary analysis was sponsored and funded by Novo Nordisk. The trial sponsor was involved in the design of the trial; the collection and analysis of data; and writing the clinical report. Dr Zinman has received grant support from Boehringer Ingelheim, AstraZeneca, and Novo Nordisk and consulting fees from AstraZeneca, Boehringer Ingelheim, Eli Lilly, Janssen, Merck, Novo Nordisk, and Sanofi. Disclosures for the coauthors are listed in the paper. Dr Pieber has received research support from Novo Nordisk and AstraZeneca (paid directly to the Medical University of Graz) and personal fees as a consultant from AstraZeneca, Bristol-Myers Squibb, Eli Lilly, Novo Nordisk, and Roche Diabetes Care. He is also the chief scientific officer of CBmed (Center for Biomarker Research in Medicine), a public-funded biomarker research company. Disclosures for the coauthors are listed in the paper. Dr Rutter reports receiving honoraria and funding to attend educational meetings from Novo Nordisk and honoraria and consulting fees from Ascensia, Cell Catapult, and Roche Diabetes Care. Dr Buse reports receiving contracted consulting fees, paid to his institution, and travel support from Novo Nordisk, Eli Lilly, GI Dynamics, Elcylex, Merck, Metavention, vTv Pharma, PhaseBio, AstraZeneca, Dance Biopharm, Sanofi, Lexicon Pharmaceuticals, Orexigen, Takeda, Adocia, Roche, NovaTarg, Shenzen High Tide, Fractyl, and Dexcom; grant support from Eli Lilly, Bristol-Myers Squibb, GI Dynamics, Merck, PhaseBio, AstraZeneca, Medtronic Minimed, Sanofi, Johnson & Johnson, Andromeda, Boehringer Ingelheim, GlaxoSmithKline, MacroGenics, Intarcia Therapeutics, Lexicon Pharmaceuticals, Scion NeuroStim, Orexigen, Takeda, Theracos, and Bayer; he also reports receiving fees and holding stock options in PhaseBio and Insulin Algorithms; he also reports serving on the board of the AstraZeneca Healthcare Foundation.

Diabetologia. Published online September 15, 2017. DEVOTE-2, DEVOTE-3, Editorial

Follow Lisa Nainggolan on Twitter: @lisanainggolan1. For more diabetes and endocrinology news, follow us on Twitter and on Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.