A total of 18,144 patients were randomized in 1:1 blinded fashion either to ezetimibe / simvastatin (n=9,067) or to simvastatin alone (n=9,077). The inclusion criteria were (a) hospitalization for ST-elevation myocardial infarction (MI), non ST-elevation myocardial infarction or unstable angina (UA) (within the previous10 days), (b) age ≥50 years, (c) at least one high risk feature such as : new ST changes on ECG, positive troponin, diabetes mellitus, prior MI, peripheral arterial disease, cerebrovascular accident, prior coronary artery bypass grafting (CABG) >3 years, and/or multivessel coronary artery disease and (d) LDL-Cholesterol levels of 50-125 mg/dL (or 50-100 mg/dL if prior lipid-lowering therapy of lese potency than simvastatin 40mg).
The primary end-point was the composite of cardiovascular (CV) death, MI, hospital admission for UA, coronary revascularization (≥30 days after randomization), or stroke. Exclusion criteria included (a) current statin treatment of potency equal or more of that of simvastatin 40mg, (b) patients in whom CABG was planned as treatment for their initial event and (c) patients on diltiazem, verapamil, amiodarone; known possible inducers/inhibitors of the CYP3A4 coenzyme.
The researchers anticipated a difference in LDL cholesterol levels of approximately 15 mg/dl, and the study after the final amendment had a 90% power to detect a ~9% relative risk reduction in the primary end-point.
The mean age of the study cohort was 64 years, there were 25% female patients, and the median follow up was 6 years. Approximately 30% had diabetes mellitus, 60% had hypertension and 20% had a history of prior MI. The acute index event on presentation was ST-elevation MI 29%, non-ST elevation MI 47% and UA 24%. Nearly 90% of the study population underwent diagnostic angiography and 70% underwent percutaneous coronary intervention. The remaining 30% was managed medically. Furthermore, the need for revascularization during follow up was approximately 30%. Up-titration to 80 mg simvastatin occurred in 27% of the simvastatin arm and 6% of the ezetimibe/simvastatin arm. Of note, baseline characteristics were fairly similar between the two arms. A discontinuation rate of 42% was observed in both arms. Withdrawn consent / lost to follow up was observed in 0.7% of the study population whereas follow up data for the primary end point was available in 91% of the study cohort.
Baseline LDL cholesterol levels were 95 mg/dl (79-110 mg/dL) in both arms; the median follow-up average was mg/dL 69.5 vs. 53.7 mg/dL (Δ -15.8 mg/dL) whereas at 1 year was 69.9 mg/dL vs. 53.2 mg/dL (Δ -16.7 mg/dL) in the simvastatin and ezetimibe / simvastatin arms, respectively. LDL cholesterol lowering was observed as early as 1 month, and appeared sustained over the duration of follow-up.
The primary endpoint of CV death/MI/UA/coronary revascularization beyond 30 days/stroke was significantly lower in the ezetimibe/simvastatin arm compared with the simvastatin arm over the duration of follow-up (7 years event rate) (32.7% vs. 34.7%, hazard ratio [HR] 0.94, 95% confidence interval [CI] 0.89-0.99; p = 0.016). This corresponded to a number needed to treat (NNT) of 50 patients to prevent one event.
Other endpoints including MI (13.1% vs. 14.8%, p = 0.002), stroke (4.2% vs. 4.8%, p = 0.05) and ischemic stroke (3.4% vs. 4.1%, p = 0.008) were all significantly lower in the ezetimibe/simvastatin arm; no differences were noted for all-cause mortality (15.4% vs. 15.3%, p = 0.78), CV mortality (6.9% vs. 6.8%, p = 0.99) and need for coronary revascularization (21.8% vs. 23.4%, p = 0.11). On subgroup analysis, patients with diabetes had a greater benefit with ezetimibe/simvastatin (HR = 0.86, p for interaction = 0.023). Moreover, albeit non-significant female and elderly patients (age≥65 years) showed also a greater benefit with the combination treatment. On-treatment analysis confirmed and further embellished the primary intention-to-treat analyses; the primary endpoint was significantly reduced in the ezetimibe / simvastatin arm compared with placebo (29.8% vs. 32.4%, HR 0.92, 95% CI 0.87-0.98, p = 0.012; NNT = 38). No differences were observed in cancer incidence (10.2% vs. 10.2%, p = 0.57), myopathy (0.2% vs. 0.1%, p = 0.32), rhabdomyolysis (0.2% vs 0.1%, p=0.37), or elevated transaminases (2.5% vs. 2.3%, p = 0.43).
The results of this study will most likely generate discussion and have an impact on treatment guidelines. However, several points regarding the study design and data analysis should be addressed to better interpreting the results. According to CTT trialists, a reduction in LDL cholesterol from 67 to 50mg/dL results in an absolute risk reduction (ARR) of 0.8% and a relative risk reduction (RRR) of 8%. A reduction in LDL cholesterol levels observed in the study, from 69.5 to 53.7 mg/dl, yielded a disproportionally high (2%) ARR and a relatively lower (5.8%) RRR. Of note, as predicted, ezetimibe use caused an additional 20% reduction in LDL cholesterol concentration. The RRR reported in the study corresponds to cumulative events during the 7 years follow up. Using the available data regarding patient-years data (total patient-years clinical follow-up = 97,822; and assuming that these were equally distributed in each arm of the study, the total patient-years for each group should be 48,911, we estimate an event rate of 5.6 events per 100 patient-years vs. 5.3 events per 100 patient-years (in the treatment arm). This absolute difference in event rate (0.3 events per 100 patient-years) represents a NNT of 333 (reduction of 1 event per year).
Interestingly, the upper 95% CI of the HR is very close to 1.00 (i.e. 0.99)! this is consistent with the finding that hard end-points such as mortality and CV mortality were not significantly different in the treatment vs control arms. The disproportionate findings regarding statistical significance reported and absolute number of events requires further consideration. The majority of the events observed in the study were coronary revascularizations (25%). All cause mortality was reported as present in 15%. These events were not statistically significant when the two arms were compared. Incidence of ischemic stroke, which was statistically significantly different between the two groups, accounted for only 4% of the events. Only MI was both statistically significantly different between the two treatments and comprised a sizeable proportion (15%) of the observed events.
Of some concern is the following:
1. In both groups, 42% of the patients discontinued the administration of the study drug.
2. Full end-point follow up data are missing in 9% of cases in both study arms. This compromises the significance of the intention to treat analysis. Again, if we assume that the 42% of subjects in the ezetimibe / simvastatin group that discontinued the study drug had an event rate similar to that in the control group (34.7%) the difference between the two study groups should be 34.7% vs 33.5% and this represents a RRR of 3.45%. Of importance, if we apply the worst clinical scenario, as recently suggested by Laufs et al. (Eur Heart J2014;35:1996–2000), corresponding to a 9% of loss of follow-up clinical data, then the RRR should be even lower.
3. A 42% discontinuation rate and the 9% loss of clinical follow up, represents a study population of approximately 50% of the initial study population. This observation weakens further the statistical significance of the ITT and also the power of the on treatment analysis.
4. An unanswered question also remains, which hypo-lipidemic therapy these patients were on after they discontinued the study drug ?
5. Unfortunately, no data are provided on adverse events such as cognitive function, dementia, neuropathies and infections.
The results of IMPROVE-IT trial indicate that in survivors of high-risk ACS, the administration of ezetimibe 10 mg on top of simvastatin 40 mg is superior to the uses of simvastatin 40 mg alone, to reduce CV events on an intention to treat analysis. The true effect and safety of the proposed combination therapy requires further investigation as findings in this study need to be taken with caution for the reasons mentioned above. IMPROVE-IT however, reaffirms the notion that “lower is better”. Hopefully, full publication of the results of the study will answer a least some of our questions.
Presented by Dr. Christopher Cannon at the American Heart Association Scientific Sessions, Chicago, IL, November 17, 2014.
Presented by Dr. Michael A. Blazing at the American Heart Association Scientific Sessions, Chicago, IL, November 18, 2014.
Laufs U, Descamps OS, Catapano AL, Packard CJ. Understanding IMPROVE-IT and the cardinal role of LDL-C lowering in CVD prevention. Eur Heart J2014;35:1996–2000
Cholesterol Treatment Trialists’ (CTT) Collaboration. Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170 000 participants in 26 randomised trials. Lancet 2010; 376: 1670–81
Cannon CP, Giugliano RP, Blazing MA, Harrington RA, Peterson JL, Sisk CM, Strony J, Musliner TA, McCabe CH, Veltri E, Braunwald E, Califf RM; IMPROVE-IT Investigators. Rationale and design of IMPROVE-IT (IMProved Reduction of Outcomes: Vytorin Efficacy International Trial): Comparison of ezetimbe/simvastatin versus simvastatin monotherapy on cardiovascular outcomes in patients with acute coronary syndromes. Am Heart J 2008;156:826-32
Blazing MA, Giugliano RP, Cannon CP, Musliner TA, Tershakovec AM, White JA, Reist C, McCagg A, Braunwald E, Califf RM. Evaluating cardiovascular event reduction with ezetimibe as an adjunct to simvastatin in 18,144 patients after acute coronary syndromes: final baseline characteristics of the IMPROVE-IT study population. Am Heart J. 2014 Aug;168(2):205-12.