REVIEW
SA JOURNAL OF DIABETES & VASCULAR DISEASE
16
VOLUME 11 NUMBER 1 • MARCH 2014
In the extension phase UKPDS study, the patients were followed
up for an additional l0 years after completion of the trial, during
which difference between HbA
1c
levels in both groups disappeared.
The follow up showed significantly reduced risk for MI in those
originally randomised to intensive glycaemic control both in the
insulin and sulfonylurea groups (RRR = 15%,
p
= 0.01) and in the
metformin group (RRR = 33%,
p
= 0.05).
3
There was also, significantly, a 13% reduction in all-cause
mortality in the intensively treated group. This persistent benefit
generated from early strict glycaemic control is known as the
legacy effect, which outlives the original reduction in HbA
1c
levels
and subsequent loss of glycaemic control. These observations are
similar to those seen in the DCCT follow-up EDIC study where
cardiovascular events, non-fatal MI, stroke and cardiovascular death
were reduced by 57% despite loss of glycaemic separation.
2
The combined UKPDS and DCCT/EDIC studies show that
• glycaemic control is important in reducing long-term macro-
vascular complications
• a very long follow up period is necessary to demonstrate any
benefit
• early glucose control is important (metabolic memory).
Medium-term effects of glycaemic control
Action to Control Cardiovascular Risk in Diabetes (ACCORD)
This landmark study was designed to determine whether the CVD
event rate could be reduced by intensively treating hyperglycaemia,
hypertension and dyslipidaemia in a double 2 × 2 factorial design.
The trial was based on the hypothesis that a 1.5% difference in
HbA
lc
level would result in a 15% difference in a population of
high-risk diabetic individuals having a 3% annual CVD event rate.
4
The study included 10 251 patients with established type 2 DM
and one-third having had a cardiovascular event. Patients were
randomised to intensive glucose therapy (targeting HbA
lc
< 6%
and achieving a level of 6.4%) or standard therapy (targeting HbA
lc
levels of 7.0–7.9% and achieving a level of 7.5%). A variety of
glucose-lowering therapies were used.
There was a non-significant trend towards reduction in the
primary outcome of the trial (a composite of non-fatal Ml, stroke
or CV death) with intensive control. However, unexpectedly, there
was higher all-cause mortality (CR: 1.22, 95% Cl: 1.01–1.46,
p
=
0.04). Higher rates of severe hypoglycaemia and weight gain were
reported in the intensive glycaemic control group. Patients with
high HbA
lc
levels at baseline had higher rates of hypoglycaemia, as
were those who did not respond properly with a fall of HbA
lc
in the
intensive-control group. The explanation for incremental mortality
remains unresolved; possible explanations include hypoglycaemia
precipitating CV death, pernicious effects of specific drugs or
combinations, and a chance finding.
The ADVANCE trial
The study was conducted to determine whether intensive glucose
lowering would reduce the risk of microvascular and macrovascular
events in individuals with type 2 DM and vascular risk factors,
compared to standard conventional cases. The study involved
11 140 subjects and the mean duration of follow up was five
years. The patients were randomised to intensive versus standard
glucose control with gliclazide plus other drugs in the intensive arm
compared with other drugs in the standard-control group.
Mean HbA
lc
level achieved was 6.5% in the intensive group
compared with 7% in the standard group. The incidence of
combined major macro- and microvascular events was significantly
reduced (HR: 0.9, 95% CI: 0.82–0.98,
p
= 0.01) in the intensive-
control group. This was primarily driven by a reduction in the
progression of albuminuria or the emergence of new nephropathy.
The CV component of the primary event was not significantly
reduced by intensive glucose control. There was no evidence of
increase in all-cause mortality. Actually there was a non-significant
trend towards reduction in all-cause mortality.
5
The VADT study
The trial included American veterans and 90% were males. A
variety of glucose-lowering agents was used, including metformin,
glimepiride, rosiglitazone and insulin. An HbA
lc
level of 6.9% was
achieved in the intensified-control arm compared with an HbA
1c
level of 8.4% in the standard-treatment arm.
After a median follow up of 6.5 years, no significant lowering
of composite CV outcomes was noted in the intensive-control
group. Severe hypoglycaemia was more prevalent in the intensive-
control group. The benefits of intensive control were apparent
only in individuals with a shorter duration of diabetes, lower HbA
lc
levels, and absence of CVD at baseline. Table 1 shows the baseline
characteristics of the ACCORD, ADVANCE and VADT trials.
6
Insights from the ACCORD, ADVANCE and VADT trials
• An important finding from all three studies is the suggestion
that a beneficial effect of intervention in glycaemic control was
more likely in association with shorter disease duration.
• In the ACCORD, study participants with baseline A
1c
levels <
8%, rather than having adverse effects of intensive glycaemic
treatment on mortality, showed a significant reduction in
primary outcome, favouring such treatment. Similarly, in the
ADVANCE trial, the combined macro- and microvascular primary
outcome benefit of interventional glycaemic control was seen
in participants without a baseline history of macrovascular
disease. Similarly, in the VADT trial, patients who had composite
outcome events had a longer duration of diabetes, higher HbA
lc
levels and more coronary arterial calcification.
• The effect of hypoglycaemia may be of importance. In the
ACCORD study, although investigators stated that this
was not a mediator of increased mortality associated with
intensive therapy, intensive interventions was associated with
significant, severe hypoglycaemia. The ADVANCE and VADT
study groups similarly have reported high incidences of severe
hypoglycaemia.
• A meta-analysis of these three trials suggests that reduction in
HbA
1c
levels of 1% is associated with a 15% of relative risk
reduction in non-fatal Ml, but without benefit on stroke or all-
cause mortality.
• The conclusion from these three trials is that intensive glycaemic
control should be appropriately applied in an individualised
manner, taking into account age, duration of diabetes and
history of CVD.
• Despite the fact that ACCORD, ADVANCE and VADT showed
no benefit of intensive glucose control on primary CV endpoints
in type 2 DM, subgroup analyses suggest that any potential
benefit on CV outcomes and mortality depends on multiple
interrelated factors, such that medications capable of exerting
direct CV therapeutic effects may be required to see a CV benefit.
