You don't need to be signed in to read BMJ Blogs, but you can register here to receive updates about other BMJ products and services via our site.

Primary Care Corner with Geoffrey Modest MD: ADVANCE trial follow-up and goal A1C

22 Jan, 15 | by EBM

By: Dr. Geoffrey Modest

​​​NEJM just released the 6 year follow-up of the of the ADVANCE trial, one of the big trials which led to a re-evaluation/relaxation of the targeted A1C goals for diabetic patients. There were 2 arms to the study, with patients randomized to intensive vs. control regimens for A1c and for blood pressure control.

The initial study (see dm ADVANCE trial nejm 2008 in dropbox, or DOI: 10.1056/NEJMoa0802987):

–​11140 patients with Type 2 diabetes randomized to standard vs. intensive control with goal A1c of 6.5%, mean age 66, 43% female, average duration of diabetes 8 years, 32% with history of macrovascular disease, 10% with microvascular, baseline A1c was 7.5% (i.e., a group with not-so-bad control to begin with). Followed 5 years.

–Patients assigned to intensive glucose control were given gliclazide modified release 30-120mg, with sequential addition/increase dose of metformin, then thiazolidinediones (TZD), then acarbose and then insulin (especially basal insulin, then adding short-acting prandial insulin) — this was the treatment protocol, though practitioners were able to do their own treatments

​–Achieved A1c was 6.5 in the intensive group and 7.3 in the control group

​–Primary endpoint: composite of major vascular events (cardiac, stroke) plus major microvascular events (new or worsening nephropathy or retinopathy) were decreased from 20.0% vs.. 18.1% by intensive treatment [HR 0.90 (0.82-0.98, p=0.01)], and, separately, major microvascular events were decreased from 10.9% to 9.4% [HR 0.86 (0.77-0.97, p=0.01)] because of reduction in nephropathy, which decreased from 5.2% to 4.1% [HR 0.79 (0.66-0.93, p=0.006)]. No significant effect on macrovascular events [HR 0.94 (0.84-1.06, p=0.32)] — i.e., the driver for the endpoint improvement was in microvascular events (not surprisingly, from other studies) with only a slight trend to improvement in macrovascular events

​–Another component of this study looked at blood pressure (see dm ADVANCE bp wing lancet 2007 in dropbox, or DOI:10.1016/S0140-6736(07)61303-8​), which found that adding perindopril 2mg plus indapamide 0.625mg as fixed combination to current blood pressure therapy led to a mean reduction of blood pressure of 5.6/2.2 mmHg (baseline 145/81), a relative risk reduction of major macrovasc or microvasc events by 9% (15.5% vs. 16.8% over 4.3 years) and similar reductions of each one of these events independently, though neither reached clinical significance. However, there were significant reductions in death from cardiovascular disease (relative risk reduction of 18% — 3.8% vs. 4.6%, with CI 0.68-0.98 and p=0.03) and death from any cause (14% — 7.3% vs. 8.5%, with CI 0.75-0.98 and p=0.03).

The 6 year follow-up study (see dm ADVANCE trial follow-up norm 2014 in dropbox, or DOI: 10.1056/NEJMoa1407963)

​–8494 people participated in the post-study follow-up, with primary endpoint of death from any cause and major macrovascular events

–No diff at time of follow-up between groups in terms of blood pressure (mean 145/81 mmHg) or A1c levels (mean 7.5%)

–The group with tighter blood pressure control in the study (those on perindopril/indapamide) continued with significant mortality benefit, though attenuated: for death from any cause [HR 0.91 (0.84-0.99, p=0.03)], and cardiovascular deaths [HR 0.88 (0.77-0.99, p=0.04)].

​–No difference in either primary endpoint for intensive diabetes control, with hazard ratios of 1.00 for each.

So, in this study tighter blood pressure control did have long-term benefit even 6 years after a 5-year intervention ended (and, of note, the blood pressure achieved during those 5 years was in the mid 130s/mid-70s, which is a tad lower than JNC-8 recommends at 140/90). The imperative seems clear that we need to treat hypertensive diabetics pretty aggressively. The point of this blog post, however, is to look at A1c goals for diabetics, which I think is still pretty unclear, and that the release of the ADVANCE and ACCORD studies have perhaps inappropriately led people to dismiss tight glucose control. I think there is still a strong argument for tight control, with a few caveats, for the following reasons:

–It is clear through many studies that tight control is important, in both types 1 and 2 diabetes, in preventing the development and progression of microvascular diabetic complications (e.g. retinopathy, neuropathy, nephropathy). And this should not be dismissed or understated.

–In terms of macrovascular complications (e.g., heart attack, stroke — the ones that tend to kill you), the data for real clinical outcome benefit from other studies are clearest for metformin

​–The ADVANCE and ACCORD studies are significantly methodologically flawed (some in hindsight, i.e. after the study was conceived and executed)

–The ADVANCE study above had an unacceptable sequence of drugs (by today’s standards), not relying on metformin as first line. And it used a very long-acting sulfonylurea (modified release gliclazide is a long-acting sulfonylurea with a half-life of 10.4 hours and a duration of action of 10-24 hours), and we no longer use long-acting sulfonylureas because of the risk of hypoglycemia and (to me) the likely increased heart disease risk of the attendant hyperinsulinemia– i.e., we use short-acting glipizide as the preferred agent

–The ACCORD trial (the only of the recent trials showing adverse effects of tighter control with achieved A1c of 6.4% vs. 7.5%, finding a significant 22% increase in mortality), in attempting to achieve better glucose control in their “intensive control” arm, used a lot of TZDs (91% vs. 58% in the control arm), the predominant one being rosiglitazone, which subsequently has been shown to increase heart disease risk as well as heart failure. Also 3.5kg weight gain in those on intensive therapy, which might have been a factor.

–And, secondary analyses of the ACCORD trial found a few really interesting things

–The major difference in mortality was explained by whether one was assigned to the intensive vs. control groups and not by the A1C achieved (suggesting that the issue was not the actual A1c achieved but the rigor in trying to achieve it — e.g., more meds with their attendant adverse reactions)

–Those who achieved lower A1c (down to an A1c of 6%) actually had lower mortality: even within the intensive group, those who achieved the lowest A1c did the best (reinforcing the interpretation that those with harder-to-treat diabetes despite flogging them with more drugs, don’t do as well with aggressive management)

–Other studies (a recent Italian one, for example — Greenfield, et al. Annals of Internal Medicine, 2010) has shown that diabetic patients with lots of medical comorbidities actually do NOT do better with tight control (i.e., no difference in cardiovascular events if A1c is <6.5 or <7.0), but those without comorbidities had 37% fewer cardiovascular events with A1c <6.5 vs. <7.0.

So, what is to be done?? To me, these studies reinforce that those patients who can easily achieve really tight control with minimal clinically significant hypoglycemia should have as tight control as we can achieve, with metformin being the first line agent. This group will typically be younger patients with shorter duration of diabetes and fewer comorbidities. On the other extreme, patients with long-standing hard-to-control diabetes who are older and have lots of comorbidities do not clearly benefit by aggressive treatments and probably should have much less tight control. What is the goal number for them? Who knows.??? There are no studies that I know of which show that A1c of 9 is much worse than 8 (and I certainly have my share of patients for whom even a goal of 9 is difficult to achieve…). So, when all else fails, treat the patient and not the number. And, in terms of priorities, there is likely much more mortality benefit from focusing on smoking cessation, lipid management, and hypertension control than lowering the A1c.

By submitting your comment you agree to adhere to these terms and conditions
You can follow any responses to this entry through the RSS 2.0 feed.
EBM blog homepage

Evidence-Based Medicine blog

Analysis and discussion of developments in Evidence-Based Medicine Visit site



Creative Comms logo

Latest from Evidence-Based Medicine

Latest from EBM