ACCORD Lipid

Published

ACCORD Study Group. "Effects of combination lipid therapy in type 2 diabetes mellitus". The New England Journal of Medicine. 2010. 362(17):1563-1574.
PubMed • Full text • PDF

Clinical Question

Among adults with T2DM on statin monotherapy, does addition of a fibrate reduce risk of CVD?

Bottom Line

Among adults with T2DM on statin monotherapy, the addition of fenofibrate does not reduce CVD endpoints.

Major Points

Adults with T2DM experience excess morbidity and mortality from CVD. Atherogenic cholesterol subtypes and inflammation are thought to be on the causal pathway of CVD. Statins promote a more optimal lipid profile through suppression of HMG-CoA reductase and reduce inflammation through their effects on multiple pathways. Multiple studies have found statins to be effective in primary and secondary prevention of CVD in various populations. Among adults with T2DM, CARDS (2004) found atorvastatin to reduce the hazard of CVD endpoints by 37%. However, CVD remained elevated at 1.54/100 person-years during the follow-up period. Whether further optimization of lipids with non-statin medications would benefit diabetics was unknown.

Fibrates appear to increase HDL and lower TGs through activation of PPAR-alpha. In a time before the widespread use of statins, the fibrate gemfibrozil was shown to have a modest effect in secondary prevention of adult men with CAD (25% of whom had diabetes) in VA-HIT (1999). Similarly, the fibrate fenofibrate reduced the incidence of some CVD outcomes in adults with T2DM in FIELD (2004), though use of statins was low in this trial.

Published in 2010, the Action to Control Cardiovascular Risk in Diabetes (ACCORD) lipid substudy evaluated add-on fenofibrate therapy to simvastatin 20-40 mg monotherapy among 5,518 adults with T2DM. (ACCORD Lipid was a substudy of ACCORD and employed 2x2 factorial design in about half of ACCORD's participants.) With 4.7 years of follow-up, there was no significant difference in the primary outcome of nonfatal MI, nonfatal stroke, or CVD mortality. Fenofibrate was generally well tolerated but was associated with an increased risk of elevated LFTs and increased creatinine.

Given these findings, practice guidelines have focused on the use of moderate- or high-intensity statins for primary prevention of CVD in diabetics.

Guidelines

ACC/AHA Cholesterol Guidelines (2013, adapted)

• Primary prevention of ASCVD in adults with DM and LDL-C 70-189 mg/dL:
  • Moderate-intensity statin use if age 40-75 (NHLBI grade A, ACC/AHA COR I, ACC/AHA LOE A)
  • High-intensity statin is reasonable if age 40-75 and 10-year ASCVD risk score ≥7.5% (NHLBI grade E, ACC/AHA COR IIa, ACC/AHA LOE B
  • If age <40 or >75 or if LDL <70 mg/dL, it's reasonable to consider initiating or continuing stain therapy (NHLBI garde E, AHA/COR iIa, ACC/AHA LOE C)

ADA DM care (2017, adapted)

• Among adults with DM, in addition to lifestyle modifications:
  • Among all ages if ASCVD, initiate high-intensity statin (grade A)
  • If age 40-75 and no additional ASCVD risk factors, consider moderate-intensity statin (grade A)
  • If age 40-75 and additional ASCVD risk factors, consider high-intensity statin (grade B)
  • If age >75 and no additional ASCVD risk factors, consider moderate-intensity statin (grade B)
  • If age >75 and additional ASCVD risk factors, consider moderate- or high-intensity statin (grade B)

Design

• Multicenter, placebo-controlled, randomized controlled trial
• N=5518
  • Fenofibrate (n=2765)
  • Placebo (n=2753)
• Setting: 77 centers in North America
• Enrollment: 2001-2005
• Mean follow-up: 4.7 years (Note: The main ACCORD trial was stopped at 3.5 years because of increased risk of CVD events with intensive glycemic control.)
• Analysis: Intention-to-treat
• Primary outcome: First nonfatal MI, nonfatal stroke, or CVD mortality

Population

Inclusion Criteria

• Age 40-79 if prior CVD, 55-79 if no prior CVD
• T2DM on >3 months stable therapy with A1c in prior 3 months in a range:
  • 7.5-11% if not on insulin on insulin ≤1 u/kg and ≤1 oral DM med, or not on insulin and ≤2 oral DM meds
  • 7.5-9% if on insulin ≤1 u/kg and 2 oral DM meds, not on insulin plus 3 oral DM meds, on insulin >1 u/kg and no oral DM meds
• High risk for CVD, defined by the following:
  • CVD: MI, stroke, revascularization of coronary, carotid, or peripheral artery, or aniina with ischemic changes
  • No CVD but high likelihood of CVD with ≥1 of the following:
    • Microalbuminuria
    • ABI <0.9
    • LVH
    • ≥50% stenosis of coronary, carotid, or peripheral artery
  • ≥2 of the following CVD risk factors:
    • Lipid-lowering medications or no lipid medis and LDL >130 mg/dL
    • HDL <40 mg/dL for men or <50 mg/dL for women
    • On antihypertensive medication or BP ≥140/≥95 mm Hg
    • Current smoker
    • BMI ≥32 kg/m²
• Cholesterols within a range: LDL 60-180 mg/dL if not on lipid med, HDL <55 mg/dL for women or blacks or <50 mg/dL for all others, TG <750 mg/dL not on therapy or <400 mg/dL on lipid med

Exclusion Criteria

• Hypoglycemic coma/seizure in prior 12 mo or hypoglycemia requiring intervention in prior 3 months with glucose <60 mg/dL
• T1DM
• Unwilling to check FSBG or administer insulin
• BMI ≥45 kg/m²
• Creatinine >1.5 mg/dL, transaminase >2x ULN
• Non-DM medications affecting glycemia (e.g., glucocorticoids)
• CVD event in prior 3 mo
• Symptomatic HF or prior NYHA class ≥III symptoms or EF <25%
• Life limiting or compliance-inducing medical issue
• Organ transplant
• Weight loss >10% in prior 6 mo
• Hypersensitivity to a study medication
• Need for a medication that interacts with a study medication
• Refusal to stop current medications
• Prior pancreatitis
• Uncontrolled thyroid disease
• Breast feeding, pregnant, trying to become pregnant, or not on birth control
• Prior myositis or myopathy
• Gallbladder disease

Baseline Characteristics

From all patients

• Demographics: Age 62 years, female 31%, white 68%, black 15%, hispanic 7%
  • Education: Less than HS 14%, HS or GED 26%, some college 33%, college or higher 27%
• PMH: CV event 36%, HF 5%, current smoker 15%, former smoker 46%
• Anthropomorphics and vitals: 95 kg, BMI 32 kg/m², BP 134/74 mm Hg
• Meds: Insulin 33%, metformin 62%, SU 52%, TZD 18%, ACE-inhibitor 54%, ARB 15%, ASA 56%, beta blocker 33%, thiazide 27%, statin 60%, any lipid-lowering medication 64%
• DM details: Duration 9 years, amputation 2%
• Labs: A1c 8.3%, FPG 176 mg/dL, K 4.5 mg/dL, creatinine 0.9 mg/dL, tchol 175 mg/dL, LDL 101 mg/dL, HDL 38 mg/dL, TG 162 mg/dL
• GFR (ml/min/1.73 m²):
  • 30-49: 3%
  • ≥50: 97%

Interventions

• This was a 2x2 factorial substudy (n=5,518) of the ACCORD study (n=10,251). All participants in ACCORD Lipid were also included in the Hgb A1c-targeted therapy of the main ACCORD study.
• In ACCORD Lipid, participants were randomized to a group:
  • Fenofibrate - Fenofibrate 160 mg/day, this became GFR-dose adjusted starting in 2004
  • Placebo
• All participants received simvastatin 20-40 mg/day based upon various protocol amendments outlined in the Section 6 of the supplemental appendix.

Outcomes

Comparisons are fenofibrate vs. placebo.

Primary Outcome

First nonfatal MI, nonfatal stroke, or CVD mortality

2.24 vs. 2.41/year (HR 0.92; 95% CI 0.79-1.08; P=0.32)

Secondary Outcomes

First nonfatal MI, nonfatal stroke, CVD mortality, or HF hospitalizations

5.35 vs. 5.64/year (HR 0.94; 95% CI 0.85-1.05; P=0.30)

Fatal coronary event, nonfatal MI, or UA

2.58 vs. 2.79/yr (HR 0.92; 95% CI 0.79-1.07; P=0.26)

Nonfatal MI

1.32 vs. 1.44/yr (HR 0.91; 95% CI 0.74-1.12; P=0.39)

Stroke

Any: 0.38 vs. 0.36/yr (HR 1.05; 95% CI 0.71-1.56; P=0.80)

Nonfatal: 0.35 vs. 0.30/yr (HR 1.17; 95% CI 0.76-1.78; P=0.48)

Mortality

All-cause: 1.47 vs. 1.61/yr (HR 0.91; 95% CI 0.75-1.10; P=0.33)

CVD: 0.72 vs. 0.83/yr (HR 0.86; 95% CI 0.66-1.12; P=0.26)

Fatal or non-fatal HF

0.90 vs. 1.09 (HR 0.82; 95% CI 0.65-1.05; P=0.10)

Additional Analyses

Change in lipids at 4 months

Figures depicting changes in lipids over time are in Figure 1 on page 1568. Statistical trends were similar at 4 and 12 months.

Tchol: Fenofibrate lower (P<0.001)

LDL: No difference (P=0.11)

HDL: Fenofibrate higher (P<0.001)

TG: Fenofibrate lower (P<0.001)

Subgroup Analysis

For the primary outcome.

Sex

Females: 9.0 vs. 6.6%

Males: 11.2 vs. 13.3%

P for interaction = 0.01

Dyslipidemia

TG ≥204 and HDL ≤34 mg/dL: 12.4 vs. 17.3%

Other combined lipid groups were not shown

P for interaction = 0.06

There were no other significant interactions for the primary outcome among prespecified subgroups, including by age, race, prior CVD, ACCORD A1c target group assignment, lipid levels, and Hgb A1c.

Adverse Events

Serious

There were no significant difference in any serious adverse events between the treatment groups.

Laboratory changes

ALT >5x ULN: 0.6% vs. 0.3% (P=0.03)

There were no differences in levels of CPK during follow-up.

Elevation in creatinine

>1.3 mg/dL in women: 28% vs. 19% (P<0.001)

>1.5 mg/dL in men: 37% vs. 18% (P<0.001)

Mean creatinine increased from 0.9 to 1.1 mg/dL in the first year in the fenofibrate group. There was no significant change in creatinine in the first year in the placebo group. Creatinine remained stable throughout follow-up for the fenofibrate group. Creatinine increased slightly during follow up in the placebo group from 0.9 to 1.0 mg/dL during follow-up. There was no difference in the amount of adults starting dialysis.

U albumin:creatinine 30-300 mg/g: 38.2% vs. 41.6% (P=0.01)

U albumin:creatinine >300 mg/g: 10.5% vs. 12.3% (P=0.04)

Criticisms

• There was heterogeneity in lipid profiles of participants. Those with high triglycerides and low HDL may have benefited.

Funding

• NHLBI at the NIH
• Authors with multiple financial disclosures
• The study medications were donated by Abbott Laboratories and Merck, who had no input in the study.