IMPROVE-IT Results in Detail: Numbers, Subgroups, and Time Course

At a glance

Parameter	Detail
Trial	IMPROVE-IT (IMProved Reduction of Outcomes: Vytorin Efficacy International Trial)
N	18,144
Population	Stabilized ACS within prior 10 days, LDL 50 to 125 mg/dL
Intervention	Ezetimibe 10 mg + simvastatin 40 mg daily
Comparator	Placebo + simvastatin 40 mg daily
Median follow-up	6.0 years
Primary endpoint	7-point composite: CV death, nonfatal MI, unstable angina requiring rehospitalization, coronary revascularization ≥30 days post-randomization, or nonfatal stroke
Key result	HR 0.936 (95% CI 0.89 to 0.99), p = 0.016; ARR 2.0 percentage points; NNT ~50 over 7 years

What the primary endpoint actually showed

The IMPROVE-IT trial enrolled patients within 10 days of an acute coronary syndrome event. By design, all participants had baseline LDL cholesterol between 50 and 125 mg/dL (or <100 mg/dL if already on lipid-lowering therapy). That narrow window was intentional: the investigators wanted to test whether pushing LDL below already-acceptable statin levels would reduce hard clinical events.

The primary composite endpoint occurred in 2,572 of 9,067 patients (32.7%) in the ezetimibe-simvastatin group versus 2,742 of 9,077 patients (34.7%) in the simvastatin-monotherapy group. The absolute risk reduction was 2.0 percentage points. The hazard ratio was 0.936 (95% CI 0.89 to 0.99, p = 0.016).

That p-value cleared the prespecified alpha of 0.0364 (adjusted for two interim analyses), so the result is statistically significant by the trial's own decision rules. The number needed to treat over the median 6-year follow-up was approximately 50.

Breaking apart the composite: individual endpoint results

Composite endpoints compress clinical information. Here is each component of the primary MACE outcome separated:

Two components drove the composite: nonfatal MI and nonfatal stroke. CV death was identical between groups (HR 1.00). This pattern matters for interpreting clinical significance. The trial reduced atherothrombotic events (plaque rupture, embolism) without altering mortality over 6 years.

Key secondary endpoints

The prespecified secondary composite of CV death, nonfatal MI, and nonfatal stroke (a tighter, 3-point MACE) showed a stronger signal:

• Ezetimibe-simvastatin: 20.4%
• Simvastatin alone: 22.2%
• HR 0.90 (95% CI 0.84 to 0.97), p = 0.003

All-cause mortality was 15.3% versus 15.4% (HR 0.99, 95% CI 0.91 to 1.07, p = 0.78). No mortality benefit was observed.

These secondary results align with the biological expectation: ezetimibe blocks intestinal cholesterol absorption, lowers circulating LDL, and reduces atherosclerotic plaque progression. That mechanism prevents ischemic events. It does not address the arrhythmic, hemodynamic, or other pathways that contribute to cardiac death.

LDL response distribution

The LDL data are central to why IMPROVE-IT reshaped treatment guidelines. At one year:

The time-averaged LDL difference was approximately 16 mg/dL. This modest absolute gap produced the 6.4% relative risk reduction, a finding that maps closely onto the Cholesterol Treatment Trialists' (CTT) Collaboration meta-regression. The CTT analysis predicts a ~22% relative reduction in major vascular events per 1.0 mmol/L (38.7 mg/dL) LDL decrease. Scaling linearly, a 16 mg/dL drop predicts roughly a 9% relative reduction; the observed 6.4% sits slightly below that line but within the confidence interval of the meta-regression.

The median versus mean LDL values in IMPROVE-IT were nearly identical (53.7 vs 53.2 in the combination arm), indicating a roughly symmetric distribution without heavy skewing from outliers.

Time course: when the curves separated

One of the most debated aspects of IMPROVE-IT was the Kaplan-Meier pattern. The event curves for the primary endpoint did not separate visibly until approximately 12 months after randomization. From year 1 through year 7, the gap widened progressively.

By landmark analysis:

• Year 1: minimal separation, consistent with the time needed for LDL lowering to slow plaque progression
• Year 3: approximately 1 percentage point absolute difference
• Year 5: approximately 1.6 percentage points
• Year 7: approximately 2.0 percentage points

This widening pattern is consistent with other lipid-lowering trials. The FOURIER trial with evolocumab showed a similar early-flat, late-diverging shape. The biology makes sense: LDL reduction stabilizes existing plaques and slows new plaque formation. That process takes months to years before it translates into fewer clinical events. Clinicians expecting a rapid benefit from non-statin LDL-lowering should set patient expectations accordingly.

Subgroup analyses that matter clinically

IMPROVE-IT prespecified 18 subgroup analyses. The treatment interaction was consistent across most subgroups (interaction p-values > 0.05), but three stood out as clinically informative:

Diabetes. Among the 4,933 patients with diabetes at baseline (27% of the cohort), the absolute risk reduction was 5.5 percentage points (40.0% vs 45.5%), and the HR was 0.86 (95% CI 0.78 to 0.94). This was the largest absolute benefit in any prespecified subgroup. The 2018 ACC/AHA cholesterol guidelines specifically cited this subgroup result when recommending ezetimibe add-on therapy for diabetic patients with ACS who do not reach LDL <70 mg/dL on maximally tolerated statin.

Age ≥75 years. Older patients (n = 2,798) showed a HR of 0.80 (95% CI 0.70 to 0.90), with an absolute risk reduction of approximately 8.7 percentage points. This is notable because elderly patients are often undertreated due to concerns about polypharmacy and perceived frailty.

Baseline LDL ≥95 mg/dL. Patients in the highest baseline LDL quartile had greater absolute benefit, consistent with the principle that higher starting LDL provides more room for risk reduction.

No subgroup showed statistically significant harm, and the forest plot showed all point estimates on the benefit side or crossing 1.0 by narrow margins.

Safety profile in numbers

Safety data from 18,144 patients over 6 years provide a large exposure base. Key findings:

The cancer signal was closely watched because the earlier SEAS trial of ezetimibe-simvastatin in aortic stenosis had raised a hypothesis about excess cancer. IMPROVE-IT, with its larger sample and longer follow-up, showed no difference whatsoever (10.2% vs 10.2%). The FDA label for ezetimibe was subsequently updated to reflect these reassuring long-term data.

What the trial was not designed to answer

IMPROVE-IT established that non-statin LDL lowering reduces cardiovascular events after ACS. It did not test:

• Primary prevention. All patients had established coronary disease. Extrapolation to low-risk populations is not supported by these data.
• Ezetimibe monotherapy. Every patient received simvastatin. The trial cannot separate the effect of ezetimibe alone from the incremental LDL reduction it provides on top of a statin.
• Head-to-head comparison with high-intensity statins. The comparator was simvastatin 40 mg, a moderate-intensity statin. Whether ezetimibe plus moderate-intensity statin outperforms high-intensity statin alone (e.g., atorvastatin 80 mg) was not tested.
• Comparison with PCSK9 inhibitors. Evolocumab and alirocumab trials came later and produced larger LDL reductions with correspondingly larger relative risk reductions.

These gaps are not criticisms of the trial design. They are boundaries that clinicians and guideline writers must respect when applying the IMPROVE-IT results to individual patients.

How guidelines absorbed these results

The 2018 ACC/AHA multisociety cholesterol guideline incorporated IMPROVE-IT findings into two specific recommendations. For patients with clinical ASCVD on maximally tolerated statin whose LDL remains ≥70 mg/dL, adding ezetimibe receives a Class IIa recommendation (Level of Evidence B-R). For very high-risk ASCVD patients (including those with recent ACS, multiple prior events, or diabetes), the threshold for add-on therapy was lowered further.

The 2019 ESC/EAS dyslipidaemia guidelines went even further, setting an LDL goal of <55 mg/dL for very high-risk patients, a target directly informed by the achieved LDL of 53.7 mg/dL in the IMPROVE-IT combination arm.

Putting the effect size in context

A 6.4% relative risk reduction and 2.0 percentage point absolute risk reduction may appear modest in isolation. Context matters. Ezetimibe is a generic, well-tolerated oral tablet with minimal monitoring requirements. The benefit accrued over years and was additive to statin therapy. For a post-ACS population already on evidence-based treatment, incremental gains are expected to be smaller than those seen in statin-vs-placebo trials from the 1990s.

The trial's true contribution was conceptual: it proved the "lower is better" hypothesis for LDL cholesterol using a non-statin agent. That principle opened the door for PCSK9 inhibitor development and the current era of aggressive lipid management in secondary prevention.

Frequently asked questions

›What was the exact primary endpoint result in IMPROVE-IT?

The primary 7-point MACE composite occurred in 32.7% of the ezetimibe-simvastatin group versus 34.7% of the simvastatin-alone group, yielding a hazard ratio of 0.936 (95% CI 0.89 to 0.99, p = 0.016) over a median follow-up of 6.0 years.

›Did IMPROVE-IT show a mortality benefit for ezetimibe?

No. Cardiovascular death was 6.8% versus 6.9% (HR 1.00), and all-cause mortality was 15.3% versus 15.4% (HR 0.99). The benefit was driven entirely by reductions in nonfatal MI and nonfatal stroke.

›How low did LDL go in the ezetimibe-simvastatin arm?

The median LDL at one year was 53.7 mg/dL (IQR 38.2 to 67.1) in the combination group, compared with 69.5 mg/dL (IQR 55.1 to 84.8) in the simvastatin-alone group.

›Which subgroups benefited most from ezetimibe in IMPROVE-IT?

Patients with diabetes had the largest absolute benefit (5.5 percentage point reduction, HR 0.86). Patients aged 75 and older also showed a pronounced response (HR 0.80). Those with higher baseline LDL had proportionally greater absolute risk reductions.

›When did the Kaplan-Meier curves start to separate?

Visible separation began around 12 months. The gap widened progressively through year 7, consistent with the slow biological process of plaque stabilization from sustained LDL lowering.

›Did IMPROVE-IT find any cancer signal with ezetimibe?

No. Cancer incidence was identical at 10.2% in both groups. This resolved concerns raised by the earlier SEAS trial and led the FDA to update the ezetimibe label with reassuring long-term safety data.

›What was the number needed to treat in IMPROVE-IT?

Approximately 50 patients needed to be treated with ezetimibe-simvastatin (instead of simvastatin alone) for 6 to 7 years to prevent one primary composite event.

›How did IMPROVE-IT change cholesterol treatment guidelines?

The 2018 ACC/AHA guidelines gave ezetimibe a Class IIa recommendation as add-on therapy for ASCVD patients not at LDL goal on maximally tolerated statin. The 2019 ESC/EAS guidelines set a new LDL target of <55 mg/dL for very high-risk patients, directly informed by the 53.7 mg/dL achieved in IMPROVE-IT.

›Was IMPROVE-IT a comparison against high-intensity statins?

No. The comparator was simvastatin 40 mg (moderate intensity). The trial tested the value of adding ezetimibe to that regimen, not whether ezetimibe plus moderate statin beats high-intensity statin alone.

›How does the IMPROVE-IT effect size compare to PCSK9 inhibitor trials?

IMPROVE-IT achieved a ~16 mg/dL LDL reduction with a 6.4% relative risk reduction. PCSK9 inhibitor trials like FOURIER produced ~60 mg/dL reductions with ~15% relative reductions. The relationship between LDL lowering and event reduction is roughly proportional across both drug classes.

References

• Cannon CP, Blazing MA, Giugliano RP, et al. Ezetimibe Added to Statin Therapy after Acute Coronary Syndromes. N Engl J Med. 2015;372(25):2387-2397. PubMed
• Cholesterol Treatment Trialists' (CTT) Collaboration. Efficacy and safety of LDL-lowering therapy among men and women: meta-analysis of individual data from 174,000 participants. Lancet. 2015;385(9976):1397-1405. PubMed
• Sabatine MS, Giugliano RP, Keech AC, et al. Evolocumab and Clinical Outcomes in Patients with Cardiovascular Disease. N Engl J Med. 2017;376(18):1713-1722. PubMed
• Grundy SM, Stone NJ, Bailey AL, et al. 2018 AHA/ACC Guideline on the Management of Blood Cholesterol. J Am Coll Cardiol. 2019;73(24):e285-e350. PubMed
• Mach F, Baigent C, Catapano AL, et al. 2019 ESC/EAS Guidelines for the management of dyslipidaemias. Eur Heart J. 2020;41(1):111-188. PubMed
• Ezetimibe (Zetia) prescribing information. FDA. Label