The reduction of low-density lipoprotein cholesterol (LDL-C) levels is one of the most well-proven interventions for reducing cardiovascular risk and has become a globally implemented cornerstone of treatment for the prevention of cardiovascular disease. The lowering of serum LDL-C has been shown to not only slow the progression of atherosclerosis as assessed by anatomical surrogates such as carotid ultrasound and coronary intravascular ultrasound, but more importantly also to decrease the incidence of cardiovascular events and mortality.1,2 The most effective and widely used LDL-C-lowering agents are the 3-hydroxyl-3- methylglutaryl coenzyme A reductase (HMG-CoAse) inhibitors, better known as statins, which function by inhibition of hepatic cholesterol synthesis and upregulation of LDL-C receptors.1 More aggressive LDL-C reduction by either higher doses of the same statin or use of more efficacious statins has been demonstrated in placebo-controlled, prospective, randomised trials to achieve further reductions in the rates of cardiovascular disease.2–7
Small but significantly greater reductions in LDL-C can also be achieved at the highest approved doses of the three most used statins (simvastatin, atorvastin and rosuvastatin), with 80mg of simvastatin being equivalent to 40mg of atorvastatin and 80mg of atorvastatin being equivalent to 20mg of rosuvastatin.8 However, within the dose range of any statin, the highest approved dose offers only a limited additional LDL-C reduction of approximately 6% from the second highest dose, but at the expense of an increased incidence of a number of adverse effects.9–12 Furthermore, there are many high-risk patients in whom adequate LDL-C levels cannot be achieved with statins alone, not even with rosuvastatin 40mg.13 Such patients are prescribed a second agent that complements the action of statins, such as niacin, bile acid sequestrants (BAS) or cholesterol absorption transport inhibitors (CATi), in order to achieve their therapeutic goals. Ezetimibe, a CATi, binds to the Niemann-Pick C1-like 1 (NCP1L1) protein, selectively inhibiting the absorption of biliary and dietary cholesterol at the intestinal wall.14–16 Studies have shown that ezetimibe produces a reduction in LDL-C levels of approximately 18%, irrespective of whether it is given as monotherapy or as an adjunct to existing stable doses of any other statin.17,18 It also produces a small but significant increase in high-density lipoprotein cholesterol (HDL-C) and reductions in triglycerides and apolipoprotein B.19–21 Studies have also shown an approximate doubling in the reduction of the inflammatory marker high-sensitivity C-reactive protein (hsCRP) when ezetimibe is co-administered with a statin compared with statins alone.22
The neutral results from the recent Ezetimibe and Simvastatin in Familial Hypercholesterolemia Enhances Atherosclerosis Regression (ENHANCE) trial23 have led to some questioning of the cardiovascular benefits of ezetimibe and the LDL-C hypothesis in general. This article will review the issues raised by the ENHANCE trial and the clinical implications for future studies and LDL-C-lowering therapy.
The ENHANCE Trial – The Data
The ENHANCE study was designed to evaluate the effects of simvastatin alone or in combination with ezetimibe in 720 patients with heterozygous familial hypercholesterolaemia (FH) and an LDL-C level of at least 210mg/dl following a six-week ‘wash-out’ of all lipid-lowering drugs.24 A number of previous epidemiological studies in the general population showed a strong correlation between carotid intima-media thickness (cIMT) and cardiovascular events, as well as correlations between cIMT and a number of well-established risk factors for cardiovascular disease such as age, sex, LDL-C, blood pressure and diabetes.25 The design of ENHANCE was based on that of the Atorvastatin versus Simvastatin on Atherosclerosis Progression (ASAP), a smaller trial carried out in a similar FH population in Holland in the mid- 1990s that compared the effects of simvastatin 40mg versus atorvastatin 80mg on atherosclerotic progression as measured by cIMT.26,27 As in ASAP, the primary outcome in ENHANCE was the change in mean cIMT from baseline as measured by ultrasonography, defined as a composite measure of the mean of the thicknesses in the far walls of the right and left common carotid arteries, carotid bulbs and internal carotid arteries. Secondary outcomes included the change in the mean maximal cIMT (the thickest of the six baseline measurements) from baseline, changes in the average mean IMT of the six carotid sites and the common femoral arteries from baseline and the proportion of patients who developed new carotid-artery plaques exceeding 1.3mm.
After 24 months of daily treatment there was no significant difference (p=0.29) in the primary outcome measure: the change in mean cIMT from baseline was 0.0058±0.0037 in the simvastatin-only group and 0.0111±0.0038 in the group receiving the ezetimibe/simvastatin combination.23 Secondary outcomes were also similar between groups: mean cIMT regression was 44.4% in the simvastatin group and 45.3% in the combination group (p=0.92); changes in the mean maximum cIMT were non-significant, with an increase of 0.0103±0.0049mm in the simvastatin group and 0.0175±0.0049mm in the combination group (p=0.27); and new plaque formation was seen in 2.8% of the simvastatin group and 4.7% of the combination group (p=0.20).
There were no significant differences in cardiovascular outcomes between the two groups. The additional significant LDL-C reduction seen in the ezetimibe plus simvastatin group of 16.5% (p<0.01) was as anticipated: mean LDL-C levels fell from 317.8±66.1 to 192.7±60.3mg/dl in patients treated with simvastatin, and from 319.0±65.0 to 141.3±52.6 mg/dl in patients receiving ezetimibe/simvastatin. Furthermore, patients receiving the combination therapy experienced significantly greater reductions in triglyceride and hsCRP levels compared with the simvastatin group (p<0.01 for both). The results of ENHANCE were also different from those of ASAP in the simvastatin monotherapy group in that cIMT was not significantly different from baseline and failed to show the progression seen in ASAP. In light of the lack of regression in cIMT in the ezetimibe-treated group, the only possible way left to show ‘benefit’ resulted in questions about whether there exist any additional cardiovascular benefits with combination therapy.
The Design of the ENHANCE Trial
There has been much speculation, albeit by a vocal few, on the neutral outcome of the study, with the results being interpreted as the LDL-C reduction with ezetimibe having no additional benefit, possibly even raising the risk of cardiovascular disease, or as evidence against the LDL-C hypothesis. However, as this discussion will outline, the study itself contained a number of critical flaws in its design.
As outlined earlier, ENHANCE was based on ASAP, a smaller study that found that atorvastatin could significantly reduce the progression of cIMT in patients with FH compared with low doses of simvastatin.26,27 The rationale for selecting a similar population of patients with FH for ENHANCE was that these patients would have significantly lipid-enriched and thickened carotid arteries, reflected by a baseline cIMT that would be higher than average, as well as an accelerated progression rate even on high-dose therapy.24 This hypothesis turned out to be significantly flawed: baseline thickness was not only not thickened but also consistent with the thinnest cIMT seen in epidemiological studies in low-risk women,25 and this was further compounded by the negligible progression rate observed in both treatment groups. Where other trials studying cIMT progression had set a minimum cIMT entry criterion, ENHANCE did not. The baseline cIMT was expected to be similar to the 0.93mm observed in ASAP, but the resulting baseline measurement of 0.69mm was significantly lower than that observed in other trials, many of which list a thickness of >0.7mm as a minimum enrolment criterion.
The one-year Carotid Atorvastatin Study in Hyperlipidemic Postmenopausal Women: a Randomized Evaluation (CASHMERE), which enrolled patients with similarly thin cIMT, found no statistical difference in cIMT results when comparing placebo with atorvastatin 80mg in 399 post-menopausal women; in fact, atorvastatin resulted in more progression than placebo.28 Baseline cIMT was reported as 0.69mm, and the most likely reason for the observed ‘failure’ of CASHMERE was the lack of carotid atherosclerosis at entry or the development of the disease during the trial. Just as in ENHANCE, the data from CASHMERE point towards a flawed design, even with a valid surrogate marker, rather than flaws in the efficacy of the therapy. In comparison, a recent cIMT trial studying patients with relatively mild hypercholesterolaemia, Rosuvastatin on Progression of cIMT of Low-Risk Individuals with Subclinical Atherosclerosis (METEOR), enrolled only statin-naïve patients with a minimum mean cIMT of >0.7mm.29 Thus, the majority of patients enrolled in ENHANCE and CASHMERE, with only minimal carotid atherosclerosis, would not have qualified for a trial such as METEOR.
Recruitment into ASAP took place in 1997, when general statin use, mostly at low doses, had only recently been introduced, the highest dosage of simvastatin was 40mg and atorvastatin had just been approved for general use. Thus, in ASAP few patients had received prior statin therapy, and none of them would have received the long-term aggressive statin therapy that had become more conventional by the time the ENHANCE trial began enrolment. Since 80% of the patients enrolled in ENHANCE had previous treatment with statins and other lipid-lowering agents, and given the thin baseline cIMT indicating minimal atherosclerosis, the aggressive LDL-C reduction with the addition of ezetimibe to high-dose simvastatin may not have had any ability to further alter cIMT. Indeed, a follow-up study of the initial treatment period in ASAP found that there was minimal subsequent progression and no regression (0.005mm/year) with atorvastatin 80mg/day, suggesting that continued LDL-C lowering may have had only a minimal impact on cIMT progression after a period of statin treatment.27 This is further supported by two other FH cIMT trials where patients who continued on aggressive statin therapy showed no change and no regression in cIMT over two years: Rating Atherosclerotic Disease Change by Imaging with a new CETP Inhibitor Trial (RADIANCE 1) and Efficacy and Safety of the ACAT Inhibitor CS-505 (CAPTIVATE).30,31 The minimal atherosclerosis seen at baseline by cIMT in ENHANCE is thus very likely due to the high incidence of previous statin treatment and may therefore have led to a lack of difference in cIMT progression over 24 months of treatment with ezetimibe, in spite of the 16.5% greater reduction in LDL-C levels.
Critics of the ENHANCE study have been quick to accuse ezetimibe of having no impact on atherosclerosis or to suggest that the LDL-C hypothesis is flawed, questioning the role of LDL-C as a proven surrogate for atherosclerosis.32–34 Some critics have proposed that the inability of ezetimibe to inhibit atherosclerosis may be due to an off-target effect that negates the benefits of LDL-C lowering, comparing the function of ezetimibe with that of oral oestrogen and torcetrapib. While oral oestrogen and torcetrapib can lower LDL-C, randomised, controlled, prospective outcome trials have shown that these drugs increase the risk of cardiovascular events.35,36 The increase in cardiovascular events observed with these two agents is large and robust, and can be explained by well-documented mechanisms: oral oestrogen and progesterone are prothrombotic,37 and torcetrapib substantially raises blood pressure through increased aldosterone production.38
In comparison, the pathway by which ezetimibe lowers plasma LDL-C is similar to that of statins and BAS, where the LDL receptor is upregulated secondary to hepatic intracellular cholesterol depletion. Statins function by inhibiting cholesterol synthesis, lowering hepatic intracellular cholesterol, upregulating LDL receptors and promoting LDL-C clearance. BAS interfere with the reabsorption of bile acid in the ileum, decreasing intracellular hepatic cholesterol and upregulating LDL receptors. Ezetimibe works in a similar fashion, decreasing cholesterol return to the liver to lower hepatic intracellular levels and upregulate LDL receptors.39 Animal studies have shown that ezetimibe can reduce atherosclerosis40 and enhance macrophage efflux of cholesterol, potentially increasing reverse cholesterol transport.41 Thus, ezetimibe is unlikely to have an off-target effect.
The LDL-C hypothesis has also been called into question, where opponents believe that the pleiotropic effects of statins in reducing inflammation ultimately have a favourabe impact on atherosclerosis, and that changes in hsCRP may better predict risk reduction than decreases in LDL-C levels. While ezetimibe monotherapy induces very little reduction in hsCRP, which has led critics to question the antiatherosclerotic effects of ezetimibe relative to statins, the addition of ezetimibe to simvastatin, as in ENHANCE, results in a significantly robust incremental decrease in hsCRP, nearly doubling the effect of simvastatin alone (decrease of 23.5% versus 49.2%; p<0.01).23 The greater reduction in hsCRP is consistent with previous data that ezetimibe can accentuate this effect of statins, even in the absence of a high statin dose.22
If anything, the results from ENHANCE could raise questions about the role of hsCRP and inflammation on atherosclerosis, rather than the reliability of the LDL-C hypothesis. Data have shown a correlation between cardiovascular benefits and hsCRP reduction in lipid-lowering trials,42 although additional studies are clearly needed to determine whether this association is independent of the LDL-C reduction.
Lowering of Low-density Lipoprotein Cholesterol – Where Do We Currently Stand?
Two additional cIMT lipid-lowering trials, one with fenofibrate and another with statin alone or in combination with ezetimibe, have recently been reported. Both trials measured changes in cIMT in middle-aged patients with diabetes who were naïve to lipid-lowering therapy and lacking existing clinical evidence of cardiovascular disease: the double-blind, placebo-controlled Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) substudy and the open-label, post hoc analysed Stop Atherosclerosis in Native Diabetics (SANDS) substudy.43,44
Baseline measures of cIMT in both studies ranged between 0.8 and 1.0mm, successfully enrolling patients with significant carotid atherosclerosis. While no difference was observed in cIMT between fenofibrate and placebo after five years in the FIELD substudy (p=0.987),43 in SANDS, patients aggressively treated to a low LDL-C target of ≤70mg/dl with either high-dose statin or low-dose statin plus ezetimibe exhibited a highly significant difference (p<0.0001) in cIMT than those treated with low-dose statins to moderate LDL-C levels of ≤100mg/l.44 Rather than focusing on the actual statistical results of ENHANCE, which show no difference between treatment arms, critics have opted to deem the small numerical differences of 0.0058mm on simvastatin monotherapy versus 0.0111mm on simvastatin plus ezetimibe as potential harm.32,45 By the same illogical argument, the results from SANDS – in which the same LDL-C reduction with the ezetimibe/statin combination achieved a cIMT regression of -0.025mm compared with a regression of -0.012mm with high-dose statin monotherapy – could be interpreted as enhanced benefits for ezetimibe over statin monotherapy. Evidence from the two cIMT studies involving ezetimibe – ENHANCE and SANDS – despite the former being littered with design flaws and the latter being only a small post hoc substudy, certainly show no harm, and suggest a benefit associated with continued use of the drug, consistent with the well-proven LDL-C hypothesis. In response to any concerns that may have arisen regarding how to proceed in clinical practice, the US Food and Drug Administration (FDA) recently issued a statement following extensive review of ENHANCE, stating in part: “Based on current available data, patients should not stop taking Vytorin or other cholesterol-lowering medications.”46
After ENHANCE, critics32,45 suggested that clinicians adopt the strategy of adding to statins other drugs that have shown clinical benefits when added to statins, such as fibrates. However, in the FIELD cIMT substudy, while fenofibrate induced significant differences in LDL-C and triglyceride levels compared with placebo, it failed to show any significant difference in cIMT progress from placebo. Additionally, there is no evidence at this time that the addition of fenofibrate to statin therapy will have any cardiovascular benefit. Results from the Action to Control Cardiovascular Risk (ACCORD) study are needed to provide a more definitive answer regarding the effect of fenofibrate with statin versus statin alone on cardiovascular disease.
Indeed, current guidelines by the National Cholesterol Education Program’s Adult Treatment Panel III (ATP III) specify targets for lipoprotein target goals in light of the available clinical evidence, rather than the use of specific drugs.7,47 This approach is confirmed by the recent FDA statement.46 These guidelines are based on decades of randomised end-point trials of statins and other lipid-lowering therapies assessing clinical events that have shown that the cardiovascular risk reduction associated with a reduction in cholesterol is similar regardless of how the cholesterol lowering came about.22,48–52 The FDA and the American College of Cardiology (ACC) recommend that conclusions regarding ezetimibe not be made until larger, well-designed clinical outcome trials with ezetimibe are completed, and that ezetimibe remain an option for patients who are unable to reach their LDL-C goal on high-dose statins, who cannot tolerate statins or who can tolerate only a low-dose statin.53 The Improved Reduction of Outcomes: Vytorin Efficacy International Trial (IMPROVE-IT), treating 18,000 patients with acute coronary syndrome with simvastatin with or without ezetimibe, is currently under way, with completion planned for 2012. The study is designed to determine whether the combination of ezetimibe and simvastatin can reduce the rates of heart attacks, strokes and death.
Thus, the controversy following publication of ENHANCE data, and for that matter CASHMERE, can be largely attributed to the inadequate design of the study and the underlying lack of scientifically sound criteria for patient enrolment, rather than the study treatments themselves. It seems obvious now from ENHANCE, CASHMERE and other cIMT studies that any use of cIMT in future trials will require the use of a baseline cIMT that is consistent with the presence of detectable disease, and that patients with prior lipid-lowering therapy should be excluded from participation.
Although it appears easier to use cIMT to detect whether a drug is harmful and accelerates cIMT regardless of the baseline cIMT, it is much more difficult to show beneficial effects in the absence of a population that has met strict enrolment criteria including a proven lipid-rich atherosclerotic burden.
The ENHANCE study serves as a reminder to the preventative cardiologist or lipidologist that in spite of overwhelming evidence, the LDL-C hypothesis and ‘lower is better’ are still not accepted universally. Although statins undoubtedly have antiatherosclerotic benefits, the idea of conferring pleiotropic effects to statins over LDL-C reduction distracts many clinicians from the requisite aggressive LDL-C lowering. Thus far, LDL-C – or closely associated components of LDL, such as apolipoprotein B – remains the most specific and modifiable biomarker for reducing coronary heart disease,54 and can arguably be considered the gold standard of cardiovascular risk factors. Statin therapy should therefore remain the cornerstone for preventing cardiovascular disease, with the option of titration up to higher and more aggressive doses as needed in order to achieve the necessarily low LDL-C goals.55,56 However, considering that each doubling of a statin dose can reduce LDL-C only by an average of 6%,57 the potential for increased adverse events with higher statin doses and the fact that clinicians are reluctant to use the top dose of any statin, concomitantly administered medications such as ezetimibe, BAS and niacin are important therapies in clinical practice. Indeed, where high-dose statin is not well tolerated, a combination of ezetimibe with lower-dose statins may provide many patients with a well-tolerated and efficacious alternative to treating hyperlipidaemia.