21 October 2010
An analysis of the ACCORD study has found that intensive blood glucose control did not reduce composite advanced microvascular outcomes (renal complications, eye complications or peripheral neuropathy) compared with conventional blood glucose control. These results add to similarly negative findings from the original publication of ACCORD, in which the primary endpoint of myocardial infarction, stroke, or CV death was not reduced with intensive blood glucose control. The intensive blood glucose lowering arm of ACCORD was stopped early, after a median of 3.7 years, because of associated higher all-cause mortality.
Level of evidence:
Level 1 (good quality patient-oriented evidence) according to the SORT criteria.
Healthcare professionals should continue to follow NICE guidance on type 2 diabetes and agree individual HbA1c targets taking into account patient preference, and the balance of likely benefits and burden of treatment. A “keep it simple and safe” approach seems appropriate for the initial management of blood glucose in people with type 2 diabetes, whether aiming for macrovascular or microvascular prevention.
What is the background to this?
The glycaemia arm of the ACCORD study looked at whether intensive blood glucose control (target HbA1c <6.0%) versus conventional control (target HbA1c 7.0% to 7.9%) reduced various outcomes in type 2 diabetes. The intensive blood glucose control arm of the study was stopped early because of higher all-cause mortality (5.0% vs. 4.0%, hazard ratio [HR] 1.22, 95% confidence interval [CI] 1.01 to 1.46; number needed to harm [NNH] 95). When this arm of the study was stopped, the primary endpoint of MI, stroke or CV death, was not statistically significantly reduced with intensive blood glucose control compared with conventional control (6.9% vs. 7.2%; HR 0.90, 95%CI 0.78 to 1.04). These results were discussed in more detail in MeReC Rapid Reviews No. 64 and No. 147.
ACCORD also had predefined secondary microvascular endpoints (incidence and progression of retinopathy, nephropathy and neuropathy). These were published recently and are discussed here.
What does this study claim?
Intensive blood glucose control did not reduce the composite microvascular outcomes in ACCORD compared with conventional blood glucose control.
When the intensive blood glucose control arm of the study was stopped after a median of 3.7 years, median HbA1c was 6.3% in this arm and 7.6% in the conventional control arm. At this point, the first composite outcome of advanced renal or eye complications (dialysis or renal transplantation, high serum creatinine, retinal photocoagulation or vitrectomy) did not differ between groups (8.7% both groups; HR 1.00, 95%CI 0.88 to 1.14). The second composite outcome, which added peripheral neuropathy, did not differ significantly either (31.2% and 32.5%, respectively; HR 0.96, 95%CI 0.89 to 1.02). Intensive blood glucose control did reduce some individual disease-oriented microvascular endpoints, such as micro- and macro-albuminuria, but others, such as doubling of serum creatinine, were not. Three-line worsened visual acuity, arguably a patient-oriented outcome was improved. See table below for details.
Results were similar for the composite outcomes at study end (median 5.0 years) when patients in the intensive blood glucose lowering arm had been transferred to the conventional arm when the former was stopped.
How does this relate to other studies?
Other studies which looked at microvascular effects of intensive blood glucose control in type 2 diabetes include UKPDS, ADVANCE and VADT.
In UKPDS 33, after a median follow up of 10 years, the composite microvascular outcome (vitreous haemorrhage, retinal photocoagulation or renal failure) occurred in 8.6% of the intensive blood glucose control group treated with a sulphonylurea or insulin (median HbA1c 7.0%) and 11.4% of the conventional control group (median HbA1c 7.9%). This 25% relative reduction in risk (RR 0.75, 95%CI 0.60 to 0.93) was mainly driven by fewer cases of retinal photocoagulation, there was no statistically significant reduction in vitreous haemorrhage, blindness or renal failure. In the 10-year observational follow-up of the UKPDS study, microvascular endpoints occurred in 11.0% of the intensive control group treated with a sulphonylurea or insulin compared with 14.2% of the conventional control group (RR 0.76; 95%CI 0.64 to 0.89; P=0.001). The reduction in microvascular endpoints with metformin was not statistically significant during the study or in the 10-year observational follow up. However, both the number of patients in that arm and the number of events were potentially too small to detect a difference if one truly existed.
In ADVANCE, after a median follow up of 5 years, the microvascular endpoint of new or worsening nephropathy or retinopathy occurred in 9.4% of the intensive blood glucose control arm (mean HbA1c 6.5%) and 10.9% of the conventional blood glucose control arm (mean HbA1c 7.3%). This 14% relative reduction in risk of the composite microvascular endpoint (HR 0.86, 95%CI 0.77 to 0.97) occurred mainly because of a reduction in the development of macroalbuminuria. There was no effect on the doubling of serum creatinine and no significant effect on retinopathy.
In VADT, over a median follow up of 5.6 years, intensive blood glucose control to achieve a median HbA1c of 6.9% compared with conventional control to a median of 8.4% did not statistically significantly reduce the risk of major microvascular outcomes (retinopathy, major nephropathy or neuropathy).
The ACCORD study also had a sub-study looking at the effect of blood glucose, blood pressure and lipid control on the progression of diabetic retinopathy in 2,856 people who did not have a history of treated proliferative retinopathy at baseline. This ACCORD Eye study found intensive blood glucose control reduced the rate of progression of diabetic retinopathy but did not reduce the rates of moderate vision loss (defined as loss of visual acuity of 3 or more lines in either eye).
The microvascular results from ACCORD add to results from other studies suggesting that the effect of intensive blood glucose control on microvascular outcomes are mixed, at best.
In ACCORD, intensive blood glucose control did not reduce the advanced microvascular outcomes of end-stage renal disease, doubling of serum creatinine, severe loss of vision or vitrectomy. It did statistically significantly reduce micro- and macroalbuminuria and three-line worsened visual acuity (and cataract surgery at study end). However, there were 13 individual secondary outcomes, raising the issue that the likelihood of finding a statistically significant result by chance alone increases with the number of tests undertaken (see article by Freemantle)..
These results are similar to those from ADVANCE, where there was a reduction in macroalbuminuria, but no effect on the doubling of serum creatinine or retinopathy.
UKPDS did find a reduction in retinal photocoagulation with intensive blood glucose control. However, the UKPDS cohort was younger than the ACCORD cohort and was studied from the time of diagnosis of diabetes for a longer duration of time. As ACCORD was stopped early, it may have been too short and the power too low to observe a protective effect for advanced microvascular endpoints, which usually evolve over a longer period. The accompanying editorial points out that UKPDS took about 10 years to show efficacy of intensive blood glucose control.
It is also important to note the definitions of “intensive blood glucose control”. In UKPDS 33, the intensive arm had a median HbA1c of 7.0% over 10 years, which is higher than in the intensive arms of ACCORD (median HbA1c 6.3%) or ADVANCE (median HbA1c 6.5%).
A large, contemporary observational study done in the UK recently found that an HbA1c level of about 7.5% was associated with lowest risk of all-cause mortality in patients with type 2 diabetes. As might be expected, an increase in HbA1c above this level was associated with a greater risk of mortality, but importantly, so was a decrease below this level. This was discussed in more detail in MeReC Rapid Review No. 1017.
Healthcare professionals should continue to follow NICE guidance on type 2 diabetes and agree individual HbA1c targets taking into account patient preference, and the balance of likely benefits and burden of treatment. A “keep it simple and safe” approach seems appropriate for the initial management of blood glucose in people with type 2 diabetes whether we are hoping to prevent macrovascular or microvascular events. Reducing blood glucose to levels of about 7.5% by diet, other lifestyle measures or treatment with metformin and/or a sulphonylurea would seem optimal based on current evidence. Uncertainty remains about the benefits of reducing HbA1c from around 9.5% to 7.5% using other glucose lowering interventions when other major cardiovascular risk factors (such as smoking, blood pressure, cholesterol) are also being managed actively. Pursuing HbA1c levels below a level of 7.5% by adding insulin or another third-line drug may not confer any benefit, adds to the risks, and requires caution.
Parallel-group, RCT done in 77 clinical sites in North America.
Patients, intervention and comparison
10,251 patients with diabetes, HbA1c >7·5%, and cardiovascular disease (or ≥2 cardiovascular risk factors). 5,128 randomised to intensive blood glucose control (target HbA1c <6·0%) and 5,123 to conventional control (HbA1c 7·0–7·9%). Investigators and participants were aware of treatment group assignment.
Prespecified microvascular composite outcomes were: dialysis or renal transplantation, high serum creatinine (>291·7 micromol/L), or retinal photocoagulation or vitrectomy (first composite outcome); and peripheral neuropathy (score of >2.0 on Michigan Neuropathy Screening Instrument [MNSI]) or the first composite outcome (second composite outcome). 13 prespecified secondary measures of kidney, eye, and peripheral nerve function were also assessed.
Intensive therapy was stopped before study end because of higher mortality in that group, and patients were transitioned to standard therapy. Results at transition are given in the table below.
|First composite outcome: dialysis or renal transplant, high serum creatinine, retinal photocoagulation or vitrectomy||
1.00 (0.88 to 1.14); P=0.997
|Second composite outcome: all of first composite outcomes plus peripheral neuropathy||
0.96 (0.89 to 1.02); P=0.19
US National Institutes of Health; National Heart, Lung, and Blood Institute; National Institute of Diabetes and Digestive and Kidney Diseases; National Institute on Aging; National Eye Institute; Centers for Disease Control and Prevention; and General Clinical Research Centers.
Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you