New Studies in Glucose Control Raise Questions about Care

For clinicians who treat patients with type 2 diabetes, tight glycemic control has long been a staple of good disease management. Results from the ACCORD, ADVANCE, and VA Diabetes Trial, all presented at this year’s American Diabetes Association Annual Meeting (ADA), however, have shaken up the current thinking about intensive glucose control.

“My ‘religion’ for 30 years has been [that] the tighter the control the better,” David S.H. Bell, MB, FACE, a professor of medicine at the University of Alabama School of Medicine, Birmingham said. “For years we have believed that, based on good data, one can prevent cardiac events by excellent glycemic control, and that the lower the better.”

In fact, the notion that lowered HbA1c levels could possibly be associated with an increased risk of death goes against his basic beliefs, he said.

New findings

In February 2008, the National Heart, Lung, and Blood Institute (NHLBI) announced that it would curtail the intensive treatment arm of the North American ACCORD study. According to results presented at ADA in June, of the 10,251 patients enrolled in ACCORD, those assigned to the intensified glucose lowering arm (n=5,128) had a 22% higher relative risk for death compared with patients assigned to the standard glucose control arm — an absolute mortality increase of about 1% during 3.5 years of treatment and follow-up.

“A thorough review of the data shows that the medical treatment strategy of intensively reducing blood sugar below current clinical guidelines causes harm in these especially high-risk patients with type 2 diabetes,” NHLBI director Elizabeth G. Nabel, MD, said in a press release following the ACCORD results. “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. 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.”

David S. H. Bell

To confound matters further, shortly after the NHLBI publicized these findings, researchers from the international ADVANCE trial announced that their interim data did not reflect a difference in mortality among patients with type 2 diabetes in the trial’s high-intensity and standard treatment groups.

The ADA quickly responded, saying that it “continues to advise most people with diabetes to strive for an HbA1c of <7%, but stresses individualization of treatment goals.”

The ADA’s recommendations of HbA1c <7% has not changed since its annual meeting in June.

Clinicians and their patients have attempted to draw conclusions from these seemingly similar but significantly different trials.

Results cause commotion

Researchers launched the ACCORD trial to test the effects of three complementary medical treatment strategies on cardiovascular disease morbidity and mortality in patients with type 2 diabetes. They recruited patients with type 2 diabetes at high risk for CVD or with existing CVD to assess the effects of intensive glycemic control, of treatment to increase HDL and lower triglycerides, and of intensive blood pressure control on major CVD events. Baseline data appeared in The American Journal of Cardiology in 2007.

“We were testing whether intensively lowering blood sugar levels to target normal or near-normal levels would prevent CVD events like myocardial infarction, stroke and cardiovascular death, in patients with diabetes. [We enrolled patients] who were at especially high risk for those CVD events because they either already had CVD or had at least two additional risk factors,” Denise G. Simons-Morton, MD, PhD, the NHLBI project officer for ACCORD and a member of the ACCORD steering committee explained.

She and her fellow colleagues were targeting an HbA1c of <6% in the intensive-treatment group versus an HbA1c between 7% and 7.9% in the standard treatment group. Findings showed that targeting a lower HbA1c level was more likely to lead to a fatal event. The secondary outcome — rate of cardiovascular death due to MI, heart failure and arrhythmia — was 35% higher in the intensive group, according to results presented at ADA. Additionally, hypoglycemia and weight gain more than 10 kg were more common among intensive therapy patients.

“When we observed this higher mortality, we undertook a number of analyses of the data to try to explain why it had occurred,” William T. Friedewald, MD, said of the original study results.

Friedewald is the ACCORD steering committee chair and clinical professor of medicine and public health at Columbia University in New York.

“We looked at several subgroups: Were there people who had experienced prior heart attacks who were at an increased risk [of death]? Were patients with hypoglycemia at an increased risk? Was rosiglitazone use a factor? When we went through these analyses, there was no subgroup that had a particularly higher mortality compared with patients not in that subgroup,” he said. “All we know is that the intensive group received more intense treatment, increased insulin, and that they had to work harder at lowering their HbA1c levels. We don’t know if any or all of these factors explain the difference in mortality.”

Friedewald’s initial expectation was that as HbA1c levels decreased so would the mortality. After analyzing the data thoroughly no specific cause was found, he said, adding that the observed mortality was in fact lower than had been expected for these people out of the general community.

No increased risk

Denise G. Simons-Morton, MD, PhD, is the NHLBI project officer for ACCORD and a member of the ACCORD steering committee. Photo by Susan Dambrauskas

In a prepared statement prior to the ADA annual meeting, researchers with the ADVANCE trial announced that their interim results of 11,140 high-risk patients with type 2 diabetes provided no evidence of an increased risk of death among those patients receiving intensive treatment to lower blood glucose.

“These findings contrast with those reported [recently] by the U.S. NHBLI suggesting that intensive glucose lowering treatment levels had increased the death rate among patients with diabetes recruited to the ACCORD trial,” according to the press release.

Pavel Hamet, MD, professor of medicine, Canada Research Chair of Predictive Genomics at Université de Montréal and Canadian lead investigator and a member of the ADVANCE management committee, explained that the trial was designed to answer two questions in patients with type 2 diabetes: Does intensive treatment to lower blood pressure improve outcome? Does intensive treatment to reduce blood glucose improve outcome?

In September 2007, the ADVANCE Collaborative Group published evidence in The Lancet demonstrating that the blood pressure–lowering treatment had reduced the death rate among patients. In January, the portion of the study designed to assess the effects of the intensive treatment to reduce blood glucose was completed. As in ACCORD, this intensive treatment program was designed to lower blood glucose to levels below those usually recommended by clinical guidelines. The treatment regimen included a sulfonylurea drug, gliclazide (Diamicron MR, Servier), for all patients, as well as a range of other drugs for those not reaching target blood glucose levels.

Most recent results of ADVANCE, presented at ADA, showed that after five years of treatment with an intensive glucose lowering strategy, patients achieved a target HbA1c of <6.5%. Compared with patients who received the standard treatment, those in the intensive arm had an overall lower risk of serious diabetes complications of 10% (HR=0.90; 95% CI, 0.82-0.98), largely driven by a reduction in nephropathy of 21% (HR=0.79; 95% CI, 0.66-0.93).

“We were able to achieve an HbA1c mean level of 6.5% with remarkably few side effects. The study showed no evidence of any increased risk of death when blood glucose was intensively controlled, there was no clear increase in weight and the incidence of hypoglycemic events was remarkably low,” Stephen MacMahon, PhD, a co-principal investigator of ADVANCE, said during a session at ADA.

Where Does Intensive Glucose Control Stand Now, Given the Study Results Presented at ADA?   George Bakris The term intensive glucose control needs to be redefined in a way that intensive would include HbA1c between 6.5 and 7 rather than 6 or <6.5, or for that matter, normalization. The ADA guidelines should still be adhered to. I wasn’t surprised by the ADVANCE and VA Diabetes Trial data, but I was a little surprised by the ACCORD findings presented at the ADA. What surprised me was that the deaths were higher and, as someone who was involved with this very early on, the concern was always on nocturnal hypoglycemia, or hypoglycemia in general. In my opinion, the reason the mortality rate was higher in the low glucose control group in ACCORD was that the lower aggressive glucose control group had a much higher incidence of nocturnal hypoglycemia. This would result in substantial increases in sympathetic tone and early morning increases in BP, all of which would have contributed to higher cardiovascular risk. It had more to do with the timing of the medications and how glucose was controlled rather than the levels. Based on these findings, the message we have is that lower is not necessarily better. When we carefully look at the data, is clear that aggressive glucose control, while it does contribute to reduced mortality, it is far more important for reducing morbidity — specifically blindness, neuropathy, peripheral vascular disease, etc. The VA Diabetes Trial did not have an intensive arm, but findings did show clear reductions in morbidity. You can’t put the ADVANCE trial in the same context as ACCORD because intervention was at a much earlier time in the disease course. Obviously, aggressive BP and glucose control in that group given in early times of disease would have a far better effect. From the moment that we have a diagnosis of diabetes one should institute aggressive management to achieve the current recommended guidelines that are irrefutable, HbA1c <7, BP <140 and lipid management and LDL in the double digits. If you achieve all of those things, plus take a low dose aspirin a day, the data are very solid that this is the most you can do to reduce cardiovascular risk. For more information: George Bakris, MD, is professor of medicine and the director of the Hypertension Center at the University of Chicago-Pritzker School of Medicine.   Steven M. Haffner The current recommendations from the ADA are a goal of <7 with individual therapy perhaps <6 if it can be obtained without a lot of risk. The AACE has a goal of <6.5. The ACCORD data showed a potential for harm with intensive glucose control and neither the ADVANCE nor the VA Diabetes Trial showed benefits. Regarding the results presented at the ADA, the absence of benefit wasn’t a total surprise; on the other hand, the fact that ACCORD, ADVANCE and VA Diabetes Trial showed no benefit is a little bit discouraging. The increase in mortality in ACCORD was a real surprise; no one expected harm. The other thing that sort of stunned me was that people focused on the fact that the primary composite for CVD was down about 10% and sort of ignored that cardiovascular mortality was significantly increased. These findings are somewhat damaging for a diabetologist. One of the questions that is very interesting is what is this due to, and the answer is not completely clear. Some people believed there were signals for hypoglycemia. The ACCORD investigators said there was more hypoglycemia and it seemed to have been associated with mortality, but on the other hand they said it doesn’t explain the excess mortality. There are other important glycemic control trials that haven’t been reported that we need to keep in mind. One of them is ORIGIN, which is being conducted in new diabetics and a few IGT patients comparing insulin to placebo. If it turns out insulin is beneficial, that will be somewhat informative. The other trial that is interesting is the BARI-2 trial, which is in patients with coronary heart disease, those who seem to be most at risk in ACCORD, comparing two different treatment strategies (metformin plus rosiglitazone vs. insulin plus short-acting insulin). That may also be important because there should be a big difference in hypoglycemia between the two groups and that should inform whether hypoglycemia might be important. Current guidelines should take account all of these new research findings. My guess, however, is that the recommendations are probably not going to change, although they should have some caveats about the recommendations for lower HbA1c goals. For more information: Steven M. Haffner, MD, is professor of internal medicine at The University of Texas Health Science Center.

Experts look for answers

Ever since the ADVANCE researchers announced that their mortality data conflicted with the ACCORD results, clinicians have been forced to wonder: What factors could have possibly accounted for the difference in outcome?

“We will have to analyze all the possibilities,” Hamet said. “We will have to analyze what the patient population looks like.”

He started by describing that the ADVANCE researchers enrolled patients in Europe, Asia, Australia and Canada — not the United States.

“Canada is the only country that is the same between the ADVANCE and ACCORD trials,” he said.

Perhaps characteristics of the U.S. patient population adversely affected outcomes. However, Hamet admitted that relying on such a factor to explain these differences is likely a stretch.

“We may have recruited different patients, but they should be very similar. With type 2 diabetes, age is similar among patients and presence of complications is similar, as well,” he said.

Perhaps disparate protocols had an effect. “The treatment protocol in ADVANCE is based on one single hypoglycemic agent, and then anything else can be added to it to achieve good control,” Hamet said. “In ACCORD, it’s similar, but there is a lot more emphasis in the beginning on the intervening therapy. So there are differences in protocol.”

Simons-Morton agreed that there were various differences between the two trials.

“In ACCORD, we were testing treatment of any one of several medications available on a formulary to be used by clinicians. [The ADVANCE researchers] were testing [gliclazide] to be used as first-line treatment. So their treatments were not as variable as ours. We wanted to test the treatment strategy that would be used in clinical practice by a typical physician. … Furthermore, they were targeting an HbA1c of <6.5%, and we were targeting <6%,” she said.

According to George Bakris, MD, FASN, FAHA, director of the Hypertensive Diseases Unit at the University of Chicago-Pritzker School of Medicine, ADVANCE is a trial that got involved early to prevent progression, and ACCORD intervened to try to reduce events in patients who had already progressed.

“There was a very positive outcome in ADVANCE because the interventions were affecting known cardiovascular risk factors, and they were catching it at a time where pancreatic function was still pretty good, where blood pressure was intervened with … so there was markedly reduced risk, and as a result benefits.

“ADVANCE and ACCORD are two opposite sides of the spectrum,” Bakris continued. “On one side, you have very early disease where you are intervening and the results are gratifying. No one is really shocked about that. On the other hand, you have advanced disease where you are intervening too aggressively. The [ACCORD researchers] thought it would be better, but it was not.”

Epidemiology dictates that the higher the HbA1c, the greater the cardiovascular risk.

“That’s true, but it does not mean that reversing it necessarily reverses all that risk if you get too low,” Bakris cautioned. “That’s the point that is missed in this whole analysis.”

ADA’s Glycemic Recommendations for Adults With Diabetes HbA1c <7% Preprandial capillary plasma glucose 70-130 mg/dL Peak postprandial capillary plasma glucose <180 mg/dL Key Concepts in Setting Glycemic Goals: HbA1c is the primary target for glycemic control. Goals individualized based on the following: duration of diabetes, pregnancy status, age, comorbid conditions, hypoglycemia unawareness, individual patient considerations. More stringent glycemic goals (ie, a normal HbA1c ,6%) may further reduce complications at the cost of increased risk of hypoglycemia. Postprandial glucose may be targeted in HbA1c goals are not met, despite reaching preprandial glucose goals.

Lessons for clinical practice

The data are too early to suggest a change in the standard of care, according to Bakris. “In advanced diabetes, should you aggressively intervene to get the blood sugar towards normal? No — not because it is not necessarily good but because there are consequences to trying to get the blood sugar too low, and those consequences are not totally appreciated yet. Perhaps the ADA’s guidelines of HbA1c of >7% is really what we should be striving for and not getting it down to 6%,” Bakris said.

Friedewald hopes that through the overwhelming attention focused on the mortality outcomes, clinicians can still recognize the many positive aspects of the ACCORD trial.

“Mortality was not our primary outcome variable. The thing we were mainly looking at was a combination of mortality plus cardiovascular mortality, nonfatal MI and stroke. There was benefit having lower HbA1c, but it was not statistically significant and certainly didn’t outweigh the increased mortality, which is why we stopped the study,” he said.

“It is important that we continue to research new treatment strategies,” Hamet added. “If you look at ADVANCE, there is so much less mortality than there was in the UK Prospective Diabetes Study [UKPDS] group. Stroke is decreased, overall mortality is decreased. In the last 15 years, we have improved the morbidity and mortality of diabetes nearly threefold. That’s the good news.

“There are still things to do, there is still room for improvement, but we have really improved a very bad disease to a much less problematic disease. Let’s not just lose what we have gained.”

For more information:

• American Diabetes Association. Standards of medical care in diabetes—2008. Diabetes Care. 2008;31:S12-S54.
• Gerstein HC, Riddle MC, Kendall DM, et al. Glycemia treatment strategies in the action to control cardiovascular risk in diabetes (ACCORD) trial. Am J Cardiol. 2007;99[suppl]:34i–43i.
• Goff DC, Ismail-Beigi F, Gerstein H, et al. ACCORD trial study results. Presented at: the American Diabetes Association 68th Scientific Session; June 6-10, 2008; San Francisco.
• MacMahon S, Chalmers J, Patel AA, et al. ADVANCE trial study results. Presented at: the American Diabetes Association 68th Scientific Session; June 6-10, 2008; San Francisco.
• Patel A, MacMahon S, Chalmers J, et al. ADVANCE Collaborative Group. Effects of a fixed combination of perindopril and indapamide on macrovascular and microvascular outcomes in patients with type 2 diabetes mellitus (the ADVANCE trial): a randomised controlled trial. Lancet. 2007;370: 829-40.

Bekah Cintolo is a correspondent who primarily writes for O&P Business News’ sister publication Endocrine Today. This article originally appeared in Endocrine Today.