ACCORD

From Wiki Journal Club
Jump to navigation Jump to search
Gerstein HC, et al. "Effects of Intensive Glucose Lowering in Type 2 Diabetes". The New England Journal of Medicine. 2008. 358(24):2545-59.
PubMedFull textPDF

Clinical Question

In patients with T2DM, does intensive glycemic control targeting a HbA1c <6% versus standard glycemic control targeting a HbA1c 7-7.9% reduce the risk of CV events?

Bottom Line

In patients with T2DM, intensive glycemic control (target HbA1c <6%) increases mortality compared to standard control (target A1c 7-7.9%).

Major Points

Cardiovascular disease is a major cause of morbidity and mortality in patients with T2DM. The 2005 EDIC (an extension of the 1993 DCCT trial) demonstrated that improved glycemic control reduces the rate of long-term CVD in patients with T1DM. However, this macrovascular outcome had not been replicated in similar trials in patients with T2DM (including the 1998 UKPDS 33). Whether patients with T2DM have reduced incidence of CVD with near-euglycemic control (i.e. pushing the HbA1C goal <7%) was unclear.

The 2008 Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial randomized 10,251 patients with long-standing T2DM to either intensive (HbA1c <6%) or standard glycemic control (HbA1c 7-7.9%). After a median follow-up of 3.7 years, the trial was stopped early because intensive glycemic control was associated with increased all-cause (1.41% vs. 1.14%; P=0.04; NNH 370) and CV mortality (0.79% vs. 0.56%; P=0.02). The concurrently published ADVANCE trial found no CV benefit in a similar population, but it did not demonstrate increased mortality. Subsequently, VADT (2009) did not demonstrate a CV mortality benefit in T2DM VA patients in a similar trial design. Whether the mortality increase in ACCORD was resultant to the trial's heavy use of TZDs (i.e. rosiglitazone), which have been associated in some studies with increased CV mortality,[1] is unclear.[2]

As a result, intensive glycemic control in T2DM is only recommended for some low-risk groups.[3]

Of note, ACCORD was a double 2x2 factorial design study. Half of the patients were randomized to fenofibrate add-on to a statin medication in ACCORD Lipid and half were randomized to a BP target in ACCORD BP.

Guidelines

ADA Medical Care in DM (2013)[3]

  • Goal A1C <7% in non-pregnant adults to reduce microvascular and macrovascular disease complications (level B)
  • Goal A1C <6.5% for selected patients, provided no hypoglycemia or other adverse events (level C)
  • Goal A1C <8% for those with previous severe hypoglycemia, limited life expectancy, advanced microvascular or macrovascular complications, comorbidities, and long-standing difficult-to-control DM despite appropriate education and multiple agents including insulin (level B)

Design

  • Multicenter, double-blinded, two-by-two factorial, randomized controlled trial
  • N=10,251
    • Intensive (n=5,128)
    • Standard (n=5,123)
  • Setting: 77 North American centers
  • Enrollment: 2001-2005
  • Mean follow-up: 3.5 years (stopped after safety committee recommended discontinuation of intensive therapy)
  • Primary outcome: Annual rate of nonfatal MI or nonfatal stroke or CV death

Population

Inclusion Criteria

  • Type 2 diabetes mellitus
  • Hemoglobin A1c ≥7.5%
  • Age 40-79 years with CAD or 55-79 years with
    • Anatomical evidence of significant atherosclerosis
    • Albuminuria
    • LVH
    • ≥2 cardiovascular risk factors (dyslipidemia, HTN, current smoking, obesity)

Exclusion Criteria

  • Frequent or recent serious hypoglycemic events
  • Unwillingness to perform home glucose monitoring or insulin injections
  • BMI >45
  • Cr >1.5mg/dl
  • Serious illness

Baseline Characteristics

From the intensive arm.

  • Demographics: Age 62 years, female sex 39%, white race 64%, black race 20%, less than HS education 16%, college degree 26%
  • Diabetes detail: Duration 10 y, A1c 8.3%, fasting glucose 175 mg/dL
  • PMH: CVD 36%, HF 5%, current smoker 14%, former smoker 44%
  • Anthropometrics: BMI 32 kg/m2, BP 136/75
  • Meds: Insulin 34%, metformin 60%, SU 51%, TZD 20%, antihypertensive 85% ACE 53%, aspirin 55%, beta blocker 29%, thiazide diuretic 26%, statin 62%
  • Labs: TC 183, LDL 105, HDL in men 47, HDL men 38, TG 156, K 4.5, creatinine 0.9

Interventions

  • Randomized to intensive (targeting HbA1c <6%) or standard (HbA1c 7-7.9%) therapy

    • 46% randomized to intensive (SBP <120) vs. standard (SBP <140) blood pressure therapy
    • 54% randomized to fenofibrate vs. placebo; all received statin to achieve good LDL control
  • Intensive therapy group attended monthly visits for 4 months, then every 2 months, with additional visits and telephone calls as needed
  • Standard therapy group had glycemic control visits every 4 months

Outcomes

Comparisons are intensive vs. standard therapy.

Primary Outcome

Annual rate of nonfatal MI or nonfatal stroke or CV death
2.11% vs. 2.29% (HR 0.90; 95% CI 0.78-1.04; P=0.16)

Secondary Outcomes

Annual rate of all-cause mortality
1.41% vs. 1.14% (HR 1.22; 95% CI 1.01-1.46; P=0.04; NNH 370)
Annual rate of CV mortality
0.79% vs. 0.56% (HR 1.35; 95% CI 1.04-1.76; P=0.02)
Annual rate of nonfatal MI
1.11% vs. 1.45% (HR 0.76; 95% CI 0.62-0.92; P=0.004)
Annual rate of nonfatal stroke
0.39% vs. 0.37% (HR 1.06; 95% CI 0.75-1.50; P=0.74)
Annual rate of HF
0.90% vs. 0.75% (HR 1.18; 95% CI 0.93-1.49; P=0.17)

Additional Analyses

Adverse events
Hypoglycemia
Requiring medical assistance: 10.5% vs. 3.5% (P<0.001)
Requiring any assistance: 16.2% vs. 5.1% (P<0.001)
Heart failure: 3.0% vs. 2.4% (P=0.10)
MVC with patient as driver: 0.2% vs. 0.3% (P=0.40)
Non-hypoglycemic serious adverse event: 2.2% vs. 1.6% (P=0.03)
Fluid retention: 70.1% vs. 66.8% (P<0.001)
Clinical measures
Weight gain of >10 kg from baseline: 27.8% vs. 14.1% (P<0.001)
ALT > 3 times ULN: 1.0% vs. 1.5% (P=0.02)
LDL cholesterol: 90.8 mg/dL vs. 90.6 mg/dL (P=0.74)
Blood pressure
Systolic: 126.4 mmHg vs. 127.4 mmHg (P=0.002)
Diastolic: 66.9 mmHg vs. 67.7 mmHg (P<0.001)

Statistical analysis not presented by the authors for the following data.

Medications
Metformin: 94.7% vs. 86.9%
Secretagogue: 86.6% vs. 73.8%
Glimepiride: 78.2% vs. 67.6%
Repaglinide: 50.2% vs. 17.7%
TZD: 91.7% vs. 58.3%
Rosiglitazone: 91.2% vs. 57.5%
Alpha glucosidase inhibitor: 23.2% vs. 5.1%
Incretin: 17.8% vs. 4.9%
Exenatide: 12.1% vs. 4.0%
Insulin: 77.3% vs. 55.4%
Bolus insulin: 55.3% vs. 35.0%

Criticisms

  • Increased mortality could have been due to rosiglitazone (91% vs 57% use) or other specific agents or polypharmacy in the intensive group[2]
  • Disproportionate weight gain between the groups may have influenced the outcome[2]
  • No report of blood pressure medications used specifically ACE-inhibitors and ARBs, which confer greater benefit for diabetics than other agents[2]
  • Fewer patients in the intensive treatment arm received ACE-inhibitors (69.7% vs 71.9%)

Funding

Funding, medications, equipment, and supplies were provided by NIH grants from the NHLBI, other NIH departments, the CDC, General Clinical Research Centers, Abbott Laboratories, Amylin Pharmaceutical, AstraZeneca, Bayer HealthCare, Closer Healthcare, GlaxoSmithKline, King Pharmaceuticals, Merck, Novartis, Novo Nordisk, Omron Healthcare, Sanofi-Aventis, and Schering-Plough.

Further Reading