February 25, 2008
5 min read

ACCORD trial treatment arm halted

Intensive blood glucose treatment was stopped due to safety concerns.

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

The National Heart, Lung, and Blood Institute stopped the intensive blood glucose-lowering treatment arm of the Action to Control Cardiovascular Risk in Diabetes study after findings revealed an increased risk for death compared with a less intensive strategy, according to a National Institutes of Health press release.

The researchers enrolled 10,251 participants from 77 sites in the United States and Canada. Patients were aged 40 to 82 years and had type 2 diabetes and two or more other heart disease risk factors or had been diagnosed with heart disease.

Participants were assigned to either intensive treatment to lower their blood glucose to an HbA1c goal of ,6% (n=5,128) or standard treatment to lower their blood glucose to an HbA1c goal of 7% to 7.9% (n=5,123). The researchers also enrolled participants in one of two other ACCORD trials, examining the treatment effects on lipids or blood pressure. These two portions of the study will continue.

Patients underwent treatment for an average of four years. The intensive treatment group achieved median HbA1c values of 6.4% and the standard treatment group achieved median HbA1c values of 7.5%.

There were 257 deaths in the intensive treatment group vs. 203 deaths in the standard treatment group, a difference at study conclusion of 54 deaths, or a rate of 3 per 1,000 participants per year, according to a press release. Survival for all study participants was better than previously reported survival rates for comparable patients with type 2 diabetes, according to a Joslin Diabetes Center statement.

Participants were at especially high risk for having a myocardial infarction or stroke or for dying from cardiovascular disease, according to a press release. Specific medications or combinations of medications did not appear to be the cause of the higher death rates.

Intensive treatment group

The 10-member Data and Safety Monitoring Board recommended the intensive treatment arm of ACCORD be stopped 18 months early.

“Though we have stopped this part of the trial, we will continue to care for these participants, who now will receive the less-intensive standard treatment,” Elizabeth G. Nabel, MD, director of the Heart, Lung, and Blood Institute, said in a press release. “In addition, we will continue to monitor the health of all participants, seek the underlying causes for this finding and carry on with other important research within ACCORD.”

In a statement, the American Diabetes Association emphasized good control of blood glucose for the management of diabetes, while individualizing treatment goals. “The announcement raises concerns that one or more elements of the intensive glycemic control strategy or the level of HbA1c achieved are detrimental to some patients, but the reasons for the excess deaths are not yet known,” the ADA said in a press release. “The ADA looks forward to more analyses of these data and to other ongoing studies that may shed more light on this issue.” – by Christen Haigh

The early termination of the ACCORD trial due to excess mortality observed in the “tight” glycemic control group does not necessarily mean that glycemic control is unimportant. Glycemic control in the less-intense control group arm of the study was not as poor as in many prior trials, with about half the patients attaining HbA1c levels below 7.5%. Rather, the findings of ACCORD suggest that the primary emphasis in patients with type 2 diabetes mellitus should be on a balanced approach addressing other cardiovascular risks, notably lipids and blood pressure, as well as reasonable but not excessively “tight” glycemic control. As a matter of practicality, the findings of ACCORD may represent good news to both patients and their health providers that “normalization” of blood sugar, which is quite difficult to attain and maintain over time, is less important than moderately good glycemic control with regard to cardiovascular risk reduction, as long as other risk factors are concurrently being addressed. The findings from ACCORD justify the less-stringent glycemic target of HbA1c < 7% for most patients with type 2 diabetes mellitus.

– Stephen A. Brietzke, MD

Associate Professor of Clinical Medicine, Division of Endocrinology, Diabetes and Metabolism,

University of Missouri- Columbia School of Medicine

Regardless of which treatment arm the patients were in, their annual mortality was much lower than in previously reported series of patients with diabetes, which speaks to the risk factor modification of high blood pressure and lipid treatments. It is possible that some of the observed decline in mortality could have been due to better glycemic control, so we should not throw the baby out with the bath water. Also, we should confront the notion in ACCORD that this early, pre-publication data shows that there are actually fewer nonfatal heart attacks in a group randomized to the lower HbA1c target. However, higher case fatality rate (if a patient had a myocardial infarction and a higher risk for having fatal, unexplained, sudden cardiac death) was responsible for the excess mortality in that group.

So, we do not know the direct cause in the groups that formed the higher mortality. It does not appear to be related to the medications, and it does not appear to be directly related to a hypoglycemic episode. Why is the death rate higher if that overall event rate is lower? And that is tricky business because, as a cardiologist who knows the acute MI literature, there are mechanisms that may become operative in the setting of an acute MI or critical illness, secondary to the metabolic state of diabetes, which may be unrelated to the treatments they received or even the HbA1c. We have to keep that in mind.

The other question is, why sudden death? Sudden death usually occurs as a consequence of an acute MI. Were the sudden-death causes unexplained arrhythmias not triggered by an acute coronary syndrome? If so, one would say there is something about having a lower HbA1c that affects the threshold for life-threatening arrhythmia. In either of these outcomes, it is hard to know. Perhaps it has something to do with shifts in autonomic tone, which is the first thing that comes to mind as a cardiologist, when you have a difference in a higher case fatality rate and a higher risk for sudden death. You think of electrical disturbances of the myocardium which could relate to changes in the balance between sympathetic tone and vagal tone, which you see in patients with diabetes anyway. It could be accentuated, possibly, if they are treated to a lower HbA1c. So, there are a lot of questions here.

We should follow the recommendation that a directed attempt to get to an HbA1c of 6% is probably not something we should do. If you go to 6%, you will get to 6.5% in a clinical trial. The ACCORD patients were at 6.5%, but the researchers were actually trying to get them to 6%. In practice, going to 6.5% will probably get you to 7%. We know, in fact, that when doctors try to treat to 7%, most patients end up at 7.5% or 8%. So, I would not relax the vigor with which doctors want to lower HbA1c.

Also, doctors should realize the medications are not to blame; they seemingly were effective and not responsible for these findings. We have not overstated, but perhaps become obsessed with the cardiovascular issue in diabetes, and we have forgotten that there is disability and morbidity and mortality associated with microvascular endpoints. We know those are reduced in patients as you get tighter glycemic control. It would be nice to see something in print that one could dissect. Something is being triggered in these patients randomized to low, or lower than our current targets, and we do not know precisely what it is related to. It is possible that it something that could be treated or blocked. If that is the case, then we should treat lower, because we know that lower HbA1c will have less retinopathy, nephropathy and neuropathy. We need to find out what that mechanism is; whether or not the database coming out of ACCORD will uncover it, I do not know.

– Richard W. Nesto, MD

Lahey Clinic, Associate Professor of Medicine, Harvard Medical School