What Are the Clinical Implications of ACCORD/ADVANCE?

June 13, 2008

June 13, 2008 (San Francisco, California) — Following the reporting of a host of new data on intensive blood glucose lowering in patients with type 2 diabetes, the prevailing view of experts contacted by heartwire  seems to be that no major changes are necessary in most guidelines and that ultra-aggressive regimens to reach low blood sugar levels are not the way to go.

Most diabetes doctors seemed to be in agreement that an HbA1c level of 7% is an appropriate goal, but if patients could get down to lower levels relatively easily, without using overly aggressive drug regimens, that would probably be beneficial in terms of microvascular complications. While the current American Diabetes Association (ADA) guidelines recommend an A1c goal of 7.0%, other organizations, such as the American Association of Clinical Endocrinologists and the International Diabetes Federation, support an A1c goal of 6.5% or less.

New data from three new trials comparing intensive vs standard glucose lowering have been reported in the past week at the ADA 2008 Scientific Sessions. The Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial, which was stopped earlier this year because of an increased mortality in the intensive group, and the Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE) trial were both published in the June 12, 2008 issue of the New England Journal of Medicine [ 1, 2]. And a third trial--the Veterans Affairs Diabetes Trial (VADT)--was reported at the ADA meeting but has not yet been published.

The good news is that neither ADVANCE nor VADT has shown any sign of the increased mortality with intensive glucose lowering as seen in ACCORD, and ADVANCE also showed a reduction in microvascular renal complications in the intensive group. However, the bad news is that none of the three trials showed any macrovascular (cardiovascular) benefits of lowering blood glucose below 7, which was the major goal of the studies, which has been greeted with widespread disappointment.

The leaders of the ADVANCE and ACCORD trials seem to differ somewhat on what "bottom-line" recommendations should be made on the basis of all the new data.

ADVANCE study director Dr Anushka Patel (George Institute, Sydney, Australia) commented to heartwire : "I think we can say that if you lower blood sugar intensively in the way that ADVANCE did, you can achieve levels of 6.5 very safely, and this will give you a one-fifth reduction in renal complications. Diabetes is the most common cause of renal failure worldwide, and developing renal complications puts you at increased risk of developing future cardiovascular events as well as death from any cause. It is a clinically very important end point."

But chair of the ACCORD steering committee, Dr William Friedewald (Columbia University, New York), said: "I can't see much in the three latest studies to advocate lowering blood sugar levels much below 7. The major findings from all three trials show no benefit on cardiovascular events of going below 7. And ACCORD actually showed an increased mortality. While this may not be due to blood sugar level itself, there is still a flag that says it may be dangerous to go too low."

I can't see much in the . . . latest studies to advocate lowering blood sugar levels much below 7.

But Patel says that ACCORD and ADVANCE were two very different trials. "While the populations involved were broadly comparable, the intensive glucose control in the two studies was achieved in very different ways. The regimens used in ADVANCE on the whole were less aggressive. In ACCORD, they were more aggressive, and the use of [thiazolidinediones] TZDs and insulin was far more. They started at a higher level of A1c and dropped it down much faster than we did. We went much slower. That may be one explanation for the increased mortality in their study. But it could also be a chance finding. They had nonfatal MI going the other way," she observed.

She continued: "I'm not saying the ACCORD results were wrong, but they reflect what they did in their study. If physicians were to implement a pragmatic, flexible method of intensive glucose control like the one we advocated, we can be reassured that it will bring about benefits on kidney disease as we showed in ADVANCE. Okay, we did not show a benefit on macrovascular disease, but there was no evidence of harm."

In ADVANCE, all patients in the intensive group got gliclazide MR (a sulfonylurea) and then individual doctors could use whatever they wanted to further decrease blood sugar levels, but it was suggested that they increase gliclazide to the maximum dose and then add other oral agents, starting with metformin and acarbose, then TZDs, and then insulin. But gliclazide MR is not available in the US, making it difficult for Americans to use exactly the same strategy as that used in ADVANCE. Patel said she could not say whether substituting a different sulfonylurea would give the same result. "It is impossible to dissect the effect of any single drug in the intensive strategy, as many different drugs were used together. The only thing we can make conclusions about is the overall strategy we used," she commented.

Friedewald responds: "Yes, ADVANCE did use different drug regimens. They used a drug we didn't have, and they were not as aggressive as ACCORD. That may be the reason they didn't show harm while we did, but there is no way of knowing for sure. There are just too many variables to know."

More Microvascular Data to Come

He said he would like to see more evidence of microvascular benefits from other studies before making widespread recommendations to get everyone down to 6.5 or less. He pointed out that the microvascular results were so far available only from the ADVANCE trial, while ACCORD and VADT have not reported these yet. While many experts have said that the microvascular benefit seen in ADVANCE was no surprise, Friedewald is more cautious on this. "Although it is stated in the guidelines that reducing blood sugar will reduce microvascular complications, and everyone now thinks this has been proven, the only good evidence we have had was from the [UK Prospective Diabetes Study] UKPDS, which primarily showed a benefit just on eye disease, particularly laser therapy. They didn't find much effect on renal or neurological disease. Now we have ADVANCE, which found very little on eye or neurological disease but did show a reduction in kidney disease. We need the microvascular data from the other two trials to get a better idea what is going on here--if there really is something happening or whether we are just seeing some chance findings." He added that the ACCORD microvascular data should be ready for reporting in about six months.

Patel confirms that the ADVANCE data reported so far did not see a clear reduction in retinal events, but she pointed out that they had fewer retinal end points than expected, partly because this was driven by retinal revascularization, which is not a very sensitive end point, as it depends on whether patients had access to such therapy. "But we also took retinal photographs, which is a more sensitive and specific end point, and we haven't got that data yet," she added. These results are hoped to be reported at the European Association for the Study of Diabetes meeting in Rome in September.

Friedewald says he doesn't think the American guidelines will change as a result of the current data. "We have a lot of new information now, but I don't think the guidelines people will recommend aggressively pursuing A1c levels below 7 on the data currently available, and I think they will wait to see the results on microvascular disease from ACCORD and VADT. I certainly wouldn't advise the general public to change what they are doing right now."

He continued: "I think there's a feeling that if it's easy to get below 7 or if you can get there just by changing your diet or losing weight, that's perfectly safe and probably beneficial. If someone is newly diagnosed with diabetes and they take an oral agent and their A1c level is 6.2, that's fine. But at the moment, if someone has a level between 7 and 8, I would not recommend that they start using insulin or take many drugs to get down to below 7. Maybe I would change this advice if the microvascular data from all three studies show real benefits with lower levels; then I may advocate pushing to get them down further, but, for me, the ADVANCE data alone are not sufficient at the moment to be making such recommendations, especially given the mortality excess we saw in ACCORD. I don't think you can just ignore those data."

Whether the reduction in kidney disease we showed is worth the hypoglycemic episodes and the effort involved in getting the blood sugar down to these levels is a question we haven't answered yet.

Patel noted that European and international guidelines tend to recommend lower A1c levels than the American guidelines, going for levels lower than 6.5. "I think that is reasonable as long as you achieve blood glucose lowering in a not-too-aggressive manner," she said.

The other issues that have to be taken into consideration when advocating intensive glucose lowering are hypoglycemia and the effort and costs involved. Friedewald notes: "Hypoglycemia is always going to be a real concern when aggressively lowering blood sugar." Patel commented, "Whether the reduction in kidney disease we showed is worth the hypoglycemic episodes and the effort involved in getting the blood sugar down to these levels is a question we haven't answered yet, as we haven't done an economic analysis yet, and the intensive lowering is time and resource intensive. At the end of the day, it will come down to an individual decision for patients and doctors, and there will be considerations that come into it such as cost and the complexity involved in some of these regimens. But the strategy used in our trial can be used in clinical practice and is already commonly used."

No Changes to Guidelines Likely?

Other diabetes experts contacted by heartwire were of the opinion that current guidelines did not need to change. Professor of diabetic medicine Dr Rury Holman (Oxford University, UK) said: "My view is that current guidelines should remain as is. We should continue to strive to prevent A1c values rising as people's diabetes progresses to avoid microvascular complications but exercise caution about treating A1c values aggressively in those with longstanding diabetes and cardiovascular disease in the expectation that this will reduce CVD risk."

And Dr John Buse (University of North Carolina, Chapel Hill) who was part of the ACCORD trial and is also president of medicine and science at the ADA, took a similar view. "I think the target that makes sense for the population in general is less than 7%, but this can be individualized based on patient characteristics. I don't think we need to change the guidelines at the moment. We need to think about it at the ADA with a group of advisors, but personally I feel pretty good about the current general recommendations."

Cardiologists Disappointed

The two cardiologists who commented on the latest data for heartwire both expressed disappointment over the cardiovascular results. Dr Steven Nissen (Cleveland Clinic, OH) said: "What we have learned from ACCORD and ADVANCE is terribly disappointing, in my view--that the strategies we are using to lower blood sugar are simply not effective at reducing cardiovascular disease. Yes, the reduction in microvascular complications is important, but the magnitude of this benefit is relatively modest, considering there are more hypoglycemic episodes and other adverse effects with intensive care. I can't see a lot of benefit in struggling to get down to very low levels of A1c. I'm afraid the ADVANCE study does not advance the field much at all, in my opinion."

What we have learned from ACCORD and ADVANCE is terribly disappointing, in my view . . . The strategies we are using to lower blood sugar are simply not effective at reducing cardiovascular disease.

Dr Harlan Krumholz (Yale University School of Medicine, New Haven) voiced a similar opinion: "I certainly would not support recommending people use the strategies in the ACCORD intensive arm to get down to very low A1c levels. And to me, ADVANCE was also disappointing in that it did not show a macrovascular benefit, and the microvascular benefit was not a free ride. Maybe we should be saving the most intensive glucose lowering regimens for those at highest risk of renal complications. I think it is up to the individual patient to judge what risks they want to take. I don't think these studies have shown a substantial enough net benefit that you would say, 'Gee, everyone must do this.' But I do think they provide evidence of trade-offs and show what complex decisions patients are going to have to make to manage their diabetes."

Krumholz believes that current guidelines may still be overenthusiastic on blood sugar lowering. He told heartwire : "I think we need to be more circumspect about current recommendations. This is an example where we were so confident that blood sugar should be lowered to below 7 that we put this into the guidelines, and we were judging doctors on whether they were doing this. These were not experimental strategies that were being tested. They were mainstream therapies that we were telling people to use. But now I think we have to question whether we should be doing it after all."

It's the Strategy That's Important, Not the A1c Number

Both Krumholz and Nissen say that more attention needs to be paid to the individual strategies used to lower blood sugar rather than focusing purely on getting the A1c levels down.

Krumholz comments: "These studies drum home to me that we can't separate the idea of reducing blood sugar from the strategies that are we using to do that. Knowing that a drug lowers blood glucose obviously isn't enough. I don't think we should give up on the idea of lowering blood sugar, but the strategies used in these trials (particularly ACCORD) were not that impressive. I think what we have seen here is that the more meds you use, the more likely you are to expose patients to potential risks. We had hoped that the benefits of the glucose lowering would be worth it, but that's not how it panned out."

We can't separate the idea of reducing blood sugar from the strategies that are we using to do that.

Nissen agrees: "I have always thought simply lowering blood sugar as a goal was misguided. I think that the answer is that it really matters how you lower blood sugar. The problem with many of the drugs used to lower blood sugar is that they have they have sufficient downsides that we are not seeing cardiovascular benefit. Some of them cause increases in LDL and many cause weight gain. We need to find what drug or combination of drugs is going to produce a favorable outcome for patients that are at a high risk of heart disease. It is frustrating to be so far down this road and have no answers that are in any way statistically robust."

Nissen, who published a meta-analysis last year suggesting that rosiglitazone is associated with an increased risk of MI, is interested in determining whether this drug, which was used extensively in the ACCORD trial but not much in the ADVANCE trial, may have contributed to the increased mortality rate shown in ACCORD. But as the actual numbers of events in patients taking rosiglitazone has not been published in either paper, he says it is impossible to make a judgment on this at the moment. "The ACCORD authors say they couldn't pin the result on rosiglitazone, but the problem is that rosiglitazone was used quite extensively in both arms of the trial, so they probably won't be able to answer that question with any reliability. But it would still be helpful to know whether the MI rates were the same, higher, or lower in patients who got rosiglitazone," he said.

He added that the whole issue of which drugs to use is going to be discussed at an FDA meeting on July 1-2.

Editorials in the New England Journal of Medicine

Krumholz continues his campaign for testing individual strategies rather than targeting a specific A1c level in an editorial accompanying publication of the ACCORD and ADVANCE studies in the New England Journal of Medicine [ 3]. With coauthor Dr Thomas Lee (Partners HealthCare System, Boston, MA), he says that guidelines and performance measures should no longer support the use of targets without reference to the strategies used to achieve them. And noting that the assessment of net clinical benefit should be based on events averted or lives improved, they conclude: "We need to understand a strategy's effects on people, not just on surrogate end points."

Two other editorials in the New England Journal of Medicine also address implications of ACCORD and ADVANCE results. In one [4], Drs Robert Dluhy and Graham McMahon, both editors at the journal, summarize differences between the two trials in a table.

Differences Between the ACCORD and ADVANCE Studies

Characteristic ACCORD ADVANCE
Baseline data    
Participants, n 10 251 11 140
Mean age (y) 62 66
Duration of diabetes (median [ACCORD]/mean [ADVANCE]) (y) 10 8
Mean HbA1c at baseline (%) 8.1 7.2
History of macrovascular disease (%) 35 32
Target HbA1c value (%) <6.0 <6.5
Mean duration (y) 3.4 5.0
Treatments at study completion (intensive vs standard) (%)    
--Insulin 77 vs 55 41 vs 24
--Metformin 95 vs 87 74 vs 67
--Secretagogue (sulfonylurea or glinide) 87 vs 74 94 vs 62
--TZD 92 vs 58 17 vs 11
--Incretin 18 vs 5 Not reported
--Statin 88 vs 88 46 vs 48
--Any antihypertensive 91 vs 92 89 vs 88
--ACE inhibitor 70 vs 72 Not reported
--Aspirin 76 vs 76 57 vs 55
Outcome (intensive vs standard)    
Median HbA1c at study end (%) 6.4 vs 7.5* 6.4 vs 7.0*
Death from any cause (%) 5.0 vs 4.0* 8.9 vs 9.6
Death from cardiovascular cause (%) 2.6 vs 1.8* 4.5 vs 5.2
Nonfatal MI (%) 3.6 vs 4.6* 2.7 vs 2.8
Nonfatal stroke (%) 1.3 vs 1.2 3.8 vs 3.8
Major/severe hypoglycemia (%/y) 3.1 vs 1.0* 0.7 vs 0.4
Weight gain (kg) 3.5 vs 0.4 0.0 vs -1.0*
Current smoking (%) 10 vs 10 8 vs 8

size="1"> *The comparison between the intensive and the standard arms was significant.
Adapted from the New England Journal of Medicine. ©2008 New England Journal of Medicine.


Clinicians caring for patients with diabetes should continue to focus on smoking cessation, dietary and exercise counseling, blood-pressure control, and providing aspirin and a statin

On the cause of the excess deaths in the ACCORD trial, Dluhy and McMahon highlight that 19 of the 41 excess deaths from cardiovascular causes in the study were attributed to "unexpected or presumed cardiovascular disease," which they suggest may plausibly be related to hypoglycemia and misclassified as having a cardiovascular cause. They also point out that a subgroup of participants in the ACCORD trial gained substantial weight, but it is unclear whether these patients had higher rates of cardiovascular events or death than patients with less weight gain. Such large changes in weight could reflect an increase in body fat, which could result in increased risk, or TZD-induced sodium retention and heart failure, they note.

They conclude that the most appropriate target for A1c should remain 7, although lower individualized targets may be appropriate for the primary prevention of macrovascular disease. Noting that the lower-than-anticipated rate of cardiovascular events seen in these studies is an affirmation of the success of modern therapeutics, they make the point that "clinicians caring for patients with diabetes should continue to focus on smoking cessation, dietary and exercise counseling, blood-pressure control, and providing aspirin and a statin."

In a third editorial [ 5], Dr William Cefalu (Pennington Biomedical Research Center, Baton Rouge, LA) notes that both ACCORD and ADVANCE included patients at high risk of cardiovascular disease and that a target A1c level of approximately 7 may thus be the appropriate target in this population. But he points out that these studies did not address strategies for lowering blood sugar levels in low-risk patients who did not have cardiovascular disease or additional cardiovascular risk factors, so whether achieving glycemic targets below 7 will be beneficial to the vast majority of patients with type 2 diabetes and a low risk of cardiovascular disease remains another unanswered question.

  1. The Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med 2008; 358:2545-2559. Abstract

  2. The ADVANCE Collaborative Group. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med 2008; 358:2560-2572. Abstract

  3. Krumholz HM and Lee TH. Redefining quality--Implications of recent clinical trials. N Engl J Med 2008; 358:2537-2539. Abstract

  4. Dluhy RG and McMahon GT. Intensive glycemic control in the ACCORD and ADVANCE trials. N Engl J Med 2008; 358:2630-2633. Abstract

  5. Cefalu WT. Glycemic targets and cardiovascular disease. N Engl J Med 2008; 358: 2633-2635. Abstract

The complete contents of Heartwire , a professional news service of WebMD, can be found at www.theheart.org, a Web site for cardiovascular healthcare professionals.


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.
Post as: