We conducted a 52-week, randomized, double-blind, placebo-controlled, parallel-group, phase 3 trial. The objective was to assess the safety, side-effect profile, and efficacy of bempedoic acid therapy when it was added to the use of maximally tolerated statins for 1 year. Maximally tolerated statin therapy was defined as the highest intensity statin regimen that a patient was able to maintain, as determined by the investigator on the basis of clinical judgment and the patient’s history.

The trialprotocol (available at NEJM.org) was approved by an institutional review board or independent ethics committee at each participating institution. All the trial participants provided written informed consent.

The first author wrote the first draft of the manuscript, and all the authors had access to the data, participated in revising the manuscript, and concurred with the decision to submit the manuscript for publication. Medical writing and editorial assistance in the preparation of the manuscript was provided by JB Ashtin, a medical communications company, and funded by the sponsor, Esperion Therapeutics. The sponsor, with guidance from the steering committee, was involved in the design and conduct of the trial, in the data collection and analysis, and in the manuscript development process. All the authors vouch for the accuracy and completeness of the data and for the fidelity of the trial to the protocol.

Adults with atherosclerotic cardiovascular disease, heterozygous familial hypercholesterolemia, or both were eligible to participate if they had been taking stable doses of maximally tolerated statin therapy either alone or in combination with other lipid-lowering therapies for at least 4 weeks before screening and if they had a fasting LDL cholesterol level of at least 70 mg per deciliter (1.8 mmol per liter) during a 2-week screening period. Key exclusion criteria were the use of gemfibrozil or simvastatin at doses greater than 40 mg per day (although high-intensity atorvastatin and rosuvastatin regimens were permitted). The use of any inhibitor of proprotein convertase subtilisin–kexin type 9 (PCSK9) was prohibited starting 4 weeks before trial entry but was permitted after trial week 24 if the LDL cholesterol level was greater than 170 mg per deciliter (4.4 mmol per liter) and had increased by at least 25% from baseline. These criteria were evaluated by a staff member at the central laboratory who was not involved in the trial and who was unaware of the trial-group assignments; all others (investigators, staff, and patients) were unaware of the lipid levels (as well as the trial-group assignments and outcomes) throughout the trial (see theSupplementary Appendix).

Randomization was stratified according to the intensity of statin therapy at baseline (low, moderate, or high²) (see Section 4.3 in theSupplementary Appendix) and the presence or absence of heterozygous familial hypercholesterolemia. Patients were randomly assigned in a 2:1 ratio to receive either bempedoic acid (at a dose of 180 mg once daily) or matching placebo. Follow-up visits were conducted at weeks 4, 8, 12, 24, 36, and 52 and included the obtaining of fasting blood samples for biomarker measurement. (Laboratory analytical methods are discussed in Section 4.6 in theSupplementary Appendix.)

The primary end point of the trial was overall safety, which was assessed according to the incidence of adverse events and changes in safety laboratory variables. The severity of adverse events and their relation to the trial agent were classified according to protocol criteria by the site investigator. Reported verbatim terms were coded to preferred term and system-organ class with the use of theMedical Dictionary of Regulatory Activities (MedDRA), version 20.1. Designated clinical end points, including major adverse cardiac events, were adjudicated centrally by an independent expert committee whose members were unaware of the trial-group assignments. Monitoring of adverse events that occurred during the intervention period was conducted from the receipt of the first dose through 30 days after the receipt of the last dose of trial agent.

The principal secondary end point (principal efficacy end point) was the percentage change in the LDL cholesterol level from baseline to week 12. Additional key secondary end points were the percentage changes in the levels of non–high-density lipoprotein (non-HDL) cholesterol, total cholesterol, apolipoprotein B, and high-sensitivity C-reactive protein from baseline to week 12.

We chose a priori to use a sample of 1300 patients in the bempedoic acid group and 650 patients in the placebo group with 52-week follow-up so the trial would provide sufficient long-term exposure to the bempedoic acid program. We calculated that this sample size would allow the trial to identify an excess relative risk of 2.0 regarding adverse events occurring at rates between 1.6% and 13.6% in the placebo group (the 95% confidence interval excludes 1). We estimated that the sample size would also allow the trial to detect rare events at an incidence as low as 0.5% in the bempedoic acid group.

All the patients who received at least one dose of bempedoic acid or placebo were included in the safety analysis (safety population). All the safety data were analyzed with the use of descriptive statistics and were reported as observed, with no imputation of missing data. Efficacy analyses for the principal secondary end point were performed in the intention-to-treat population, which included all the patients who underwent randomization. Key efficacy end points were included in a step-down testing procedure to control the overall type 1 error (see Section 4.8 in theSupplementary Appendix). Missing data for these end points were imputed with the use of a pattern-mixture model. Analysis of covariance (ANCOVA) was used for the efficacy end points involving percentage change, with trial group and randomization strata as factors and with baseline value as a covariate. Because of its nonnormal distribution, the level of high-sensitivity C-reactive protein was analyzed with the use of a nonparametric analysis (Wilcoxon rank-sum test) with Hodges–Lehmann estimates and 95% confidence intervals. The percentage changes in lipid variables and biomarkers at other time points (weeks 24 and 52) and the assessment of the principal efficacy end point in subgroups of patients were analyzed with the use of ANCOVA without imputation (i.e., with observed data). Sensitivity analyses were performed for efficacy measures, including observed-data and on-treatment analyses; the on-treatment analyses excluded efficacy data that were collected more than 7 days after the last dose of trial agent.

Analyses were performed with the use of SAS software, versions 9.2 and later (SAS Institute). Full details of the statistical analysis plan are provided with the protocol.

We conducted this phase 3 trial between January 18, 2016, and February 21, 2018. A total of 3395 patients underwent screening, of whom 2230 underwent randomization at 114 sites in five countries. A total of 1488 patients were randomly assigned to receive bempedoic acid and 742 to receive placebo. Of these, 78.1% of the patients (1142 patients in the bempedoic acid group and 600 in the placebo group) completed the intervention, and 94.6% (1404 patients in the bempedoic acid group and 706 in the placebo group) continued in the trial through the week 52 visit (Fig. S2 in theSupplementary Appendix), which yielded a total of 1247.5 patient-years of exposure to bempedoic acid.

In the overall trial population, the mean age of the patients was 66.1 years. A total of 1628 patients (73.0%) were men, 2139 (95.9%) were white, 2176 (97.6%) had a history of atherosclerotic cardiovascular disease, and 79 (3.5%) had heterozygous familial hypercholesterolemia (Table 1). Regarding background use of statin therapy, among the 2230 patients in the trial, 148 (6.6%) used low-intensity therapy, 970 (43.5%) used moderate-intensity therapy, and 1112 (49.9%) used high-intensity therapy. A total of 172 patients (7.7%) were receiving ezetimibe either alone or in combination with statins, and 80 patients (3.6%) were receiving fibrates. The mean (±SD) LDL cholesterol level at baseline was 103.2±29.4 mg per deciliter (2.67±0.76 mmol per liter).

A total of 2229 patients (1487 in the bempedoic acid group and 742 in the placebo group) were included in the safety analysis. One patient in the bempedoic acid group underwent randomization in error and did not receive any dose of the medication; this patient was excluded from the safety population. The incidence of adverse events according to category and specific events (part of the primary end point of safety) is shown inTable 2. Adverse events that occurred during the intervention period, regardless of causality, were reported in 1167 patients (78.5%) receiving bempedoic acid and in 584 (78.7%) receiving placebo, with the majority of events (in 982 of 1167 patients [84.1%] and 514 of 584 [88.0%], respectively) being graded as mild to moderate. The incidence was similar when assessed according to system-organ class. The most common adverse events (occurring in >4% of the patients in either group) were nasopharyngitis, myalgia, upper respiratory tract infection, urinary tract infection, arthralgia, dizziness, muscle spasms, and diarrhea, which occurred with similar frequency in the two groups (Table S3 in theSupplementary Appendix).

The incidence of reported serious adverse events was generally small and similar in the two groups (216 patients [14.5%] in the bempedoic acid group and 104 [14.0%] in the placebo group). However, the percentage of patients who discontinued the blinded trial regimen owing to a reported adverse event was higher in the bempedoic acid group than in the placebo group (162 patients [10.9%] vs. 53 [7.1%]). The percentages of patients who discontinued the regimen owing to muscle-related adverse events were low in the two groups (31 patients [2.1%] in the bempedoic acid group and 14 [1.9%] in the placebo group).

Death occurring within 30 days after the last dose of trial agent was reported in 13 patients (0.9%) in the bempedoic acid group and in 2 (0.3%) in the placebo group. Of the 13 deaths in the bempedoic acid group, 5 were attributable to cancer (4 patients with lung cancer and 1 patient with liver metastases of unknown primary origin), which was generally diagnosed early during the course of the trial (3 cases were diagnosed within 90 days after the initiation of the trial regimen). Of the remaining deaths, 5 were due to cardiac causes in patients with an extensive history of coexisting vascular disease (2 deaths from cardiac failure, 2 from myocardial infarction, and 1 from hypertensive heart disease), and 1 each was due to sepsis after a cholecystectomy, pancreatic pseudocyst, and ischemic cerebral infarction. All 15 deaths were judged by the trial site investigator, who was unaware of the trial-group assignments, as being unlikely to be related or not related to the trial agent.

A total of 110 patients had at least one event that was centrally adjudicated as a major adverse cardiac event, with such events occurring in 68 patients (4.6%) in the bempedoic acid group and in 42 (5.7%) in the placebo group (Table 2). Adverse events that were categorized as cardiac disorders occurred in 157 patients (10.6%) in the bempedoic acid group and in 86 (11.6%) in the placebo group. The incidence of repeated and confirmed elevations in aminotransferase levels (>3 times the upper limit of the normal range) and the creatine kinase level (>5 times the upper limit of the normal range) was relatively low in the two groups (7 patients [0.5%] in the bempedoic acid group and 1 [0.1%] in the placebo group for aminotransferase levels; and in 7 [0.5%] and 1 [0.1%], respectively, for the creatine kinase level). All elevations in aminotransferase levels were transient and abated regardless of whether the trial agent was continued or discontinued.

The occurrence of muscle-related adverse events of importance (myalgia and muscular weakness) was similar in the two groups, whereas gout occurred more commonly in the bempedoic acid group than in the placebo group (18 patients [1.2%] vs. 2 [0.3%]). Uric acid levels increased slightly and the estimated glomerular filtration rate decreased slightly after the initiation of bempedoic acid therapy, but the latter increased over time (Fig. S3 in theSupplementary Appendix). The incidence of new-onset diabetes or worsening of diabetes was lower in the bempedoic acid group than in the placebo group (49 patients [3.3%] vs. 40 [5.4%]). The frequency of adverse events did not appear to differ substantially according to the intensity of background statin therapy (Table 3).

At week 12, bempedoic acid reduced the mean LDL cholesterol level by 19.2 mg per deciliter (0.50 mmol per liter). Treatment with bempedoic acid resulted in greater lowering of the LDL cholesterol level than was observed in the placebo group both at week 12 (difference, −18.1 percentage points; 95% confidence interval [CI], –20.0 to –16.1; P<0.001) and at week 24 (difference, −16.1 percentage points; 95% CI, –18.2 to –14.0; P<0.001). The differences in the changes from baseline, as compared with placebo, in the levels of non-HDL cholesterol, total cholesterol, apolipoprotein B, and high-sensitivity C-reactive protein at week 12 were –13.3 percentage points (95% CI, –15.1 to –11.6), –11.1 percentage points (95% CI, –12.5 to –9.8), –11.9 percentage points (95% CI, –13.6 to –10.2), and –21.5 percentage points (95% CI, –27.0 to –16.0), respectively (P<0.001 for all comparisons) (Table S5 in theSupplementary Appendix). Results were consistent in the on-treatment analysis (Table S6 in theSupplementary Appendix).

The majority of patients in the bempedoic acid group had a reduction in the LDL cholesterol level (Fig. S4 in theSupplementary Appendix). The effects of bempedoic acid were still apparent through week 52 (Figure 1), with minimal attenuation of effect in the on-treatment analyses over the period of 52 weeks (Fig. S5 in theSupplementary Appendix). Efficacy did not vary according to the type or intensity of background lipid-lowering therapy or across major subgroups, with the exception of there being a greater magnitude of effect with bempedoic acid therapy than with placebo among women than among men (Figure 2, and Fig. S6 in theSupplementary Appendix). Post hoc analysis showed that changes after randomization in the use of adjunctive lipid-lowering therapy (including PCSK9 inhibitors prescribed to five patients) did not appreciably affect the reduction in the LDL cholesterol level with bempedoic acid.