Skip to content
CholesterolResearch
← All studies
High tierRandomized Controlled TrialCitation verified

Ezetimibe Added to Statin Therapy after Acute Coronary Syndromes

Christopher P Cannon, Michael A Blazing, Robert P Giugliano, Eugene Braunwald, IMPROVE-IT Investigators - New England Journal of Medicine, 2015

The first trial to show that adding a NON-statin LDL-lowering drug on top of a statin further reduces events. Ezetimibe (which blocks intestinal cholesterol absorption) lowered LDL from 69.5 to 53.7 mg/dL and modestly but significantly cut the composite endpoint (HR 0.936). Because ezetimibe works nothing like a statin, this is key evidence that LDL lowering itself - not a statin side effect - drives benefit.

Key findings

Effect measures

  • Hazard Ratio: 0.936 (primary composite)95% CI 0.89-0.99
  • Absolute Risk Reduction: 2.0 percentage points (34.7% vs 32.7%)

Why this evidence tier (High)

Risk of bias:
Large double-blind randomized placebo-controlled trial run to completion.
Precision:
Very large; precise estimate, though the effect is small and the upper CI approaches 1.0.
Directness:
Hard clinical composite; directly tests non-statin LDL lowering on top of a statin.
Consistency:
Concordant with statin and PCSK9 evidence - same direction per unit LDL.
Funding / COI:
Manufacturer-funded (Merck); several authors were Merck employees.

High certainty that adding non-statin LDL lowering reduces events, strongly supporting LDL as the operative factor; the absolute benefit is modest.

Population:
18,144 patients stabilized after an acute coronary syndrome with LDL 50-100 mg/dL, randomized to simvastatin plus ezetimibe vs simvastatin plus placebo; median follow-up ~6 years.
Conflicts of interest:
Industry-funded by Merck, which manufactures ezetimibe. Several co-authors were Merck employees.
Funding:
Merck (manufacturer of ezetimibe); NCT00202878.

Limitations

  • Small absolute benefit over ~6 years; upper confidence bound near 1.0.
  • Secondary prevention (post-ACS); manufacturer-funded.

How this study is used