Effect of Serial Infusions of CER-001, a Pre-β High-Density Lipoprotein Mimetic, on Coronary Atherosclerosis in Patients Following Acute Coronary Syndromes in the CER-001 Atherosclerosis Regression Acute Coronary Syndrome Trial

Key Points

Question

Would infusing low doses of the high-density lipoprotein mimetic, CER-001, modify coronary atherosclerosis disease progression?

Findings

In this randomized clinical trial, 272 patients with an acute coronary syndrome were treated with weekly intravenous infusions of low doses of CER-001 or placebo for 10 weeks and underwent serial intravascular ultrasonography determination of coronary atheroma volume. Infusing CER-001 did not promote regression of coronary atherosclerosis compared with placebo in statin-treated patients.

Meaning

Addition of low doses of the high-density lipoprotein mimetic, CER-001, did not produce plaque regression in statin-treated patients following acute coronary syndrome.

Abstract

Importance

CER-001 is a negatively charged, engineered pre-β high-density lipoprotein (HDL) mimetic containing apolipoprotein A-I and sphingomyelin. Preliminary studies demonstrated favorable effects of CER-001 on cholesterol efflux and vascular inflammation. A post hoc reanalysis of a previously completed study of intravenous infusion of CER-001, 3 mg/k, showed that the intravenous infusion in patients with a high coronary plaque burden promoted regression as assessed by intravascular ultrasonography.

Objective

To determine the effect of infusing CER-001 on coronary atherosclerosis progression in statin-treated patients.

Design, Setting, and Participants

A double-blind, randomized, multicenter trial evaluating the effect of 10 weekly intravenous infusions of CER-001, 3 mg/kg, (n = 135) or placebo (n = 137) in patients with an acute coronary syndrome (ACS) and baseline percent atheroma volume (PAV) greater than 30% in the proximal segment of an epicardial artery by intravascular ultrasonography. The study included 34 academic and community hospitals in Australia, Hungary, the Netherlands, and the United States in patients with ACS presenting for coronary angiography. Patients were enrolled from August 15, 2015, to November 19, 2016.

Interventions

Participants were randomized to receive weekly CER-001, 3 mg/kg, or placebo for 10 weeks in addition to statins.

Main Outcomes and Measures

The primary efficacy measure was the nominal change in PAV from baseline to day 78 measured by serial intravascular ultrasonography imaging. The secondary efficacy measures were nominal change in normalized total atheroma volume and percentage of patients demonstrating plaque regression. Safety and tolerability were also evaluated.

Results

Among 293 patients (mean [SD] age, 59.8 [9.4] years; 217 men [79.8%] and 261 white race/ethnicity [96.0%]), 86 (29%) had statin prior use prior to the index ACS and 272 (92.8%) had evaluable imaging at follow-up. The placebo and CER-001 groups had similar posttreatment median levels of low-density lipoprotein cholesterol (74 mg/dL vs 79 mg/dL; P = .15) and high-density lipoprotein cholesterol (43 mg/dL vs 44 mg/dL; P = .66). The primary efficacy measure, PAV, decreased 0.41% with placebo (P = .005 compared with baseline), but not with CER-001 (−0.09%; P = .67 compared with baseline; between group differences, 0.32%; P = .15). Similar percentages of patients in the placebo and CER-001 groups demonstrated regression of PAV (57.7% vs 53.3%; P = .49). Infusions were well tolerated, with no differences in clinical and laboratory adverse events observed between treatment groups.

Conclusions and Relevance

Infusion of CER-001 did not promote regression of coronary atherosclerosis in statin-treated patients with ACS and high plaque burden.

Trial Registration

ClinicalTrials.gov Identifier: NCT2484378

This randomized clinical trial determines the effect of an intravenous infusion of CER-001 on coronary atherosclerosis progression in statin-treated patients.

Introduction

Among patients with an acute coronary syndrome (ACS), a substantial risk of recurrent ischemic cardiovascular events persists despite the use of evidence-based therapies including antiplatelet agents, statins, and coronary revascularization.¹ These observations have led to efforts to develop novel therapies for further cardiovascular risk reduction after ACS.

Based on a consistent inverse association between high-density lipoprotein (HDL) cholesterol levels and coronary heart disease (CHD) risk in epidemiologic studies^2,3 and evidence that HDL mimetic agents are atheroprotective in animal models,^4,5,6 several of these agents have been evaluated for clinical efficacy in patients with CHD.^{7,8,9,10,11,12} An initial report showed regression of coronary atherosclerosis in response to infusion of apolipoprotein A-I Milano (apoA-I) dimers.¹³ This finding stimulated further efforts to develop effective HDL mimetics.

CER-001 is a negatively charged, bioengineered pre-β HDL mimetic containing recombinant wild-type apolipoprotein A-I and sphingomyelin. The negative charge is characteristic of naturally circulating delipidated pre-β HDL, and its phospholipid composition is distinct from other HDL mimetics in clinical development. This feature has been proposed to result in enhanced lipid transport activity and favorable effects in animal models of atherosclerosis.^14,15 It also distinguishes this mimetic from others in development that contain either wild-type apolipoprotein A-I from human plasma or recombinant apoA-I Milano complexed with other phospholipid species. Initial human studies confirmed increases in cholesterol efflux capacity of plasma following infusion of CER-001,¹⁶ and imaging studies in small numbers of patients with genetic dyslipidemia reported beneficial effects on plaque burden and inflammatory activity in the aorta and carotid arteries.^17,18

The Can HDL Infusions Significantly Quicken Atherosclerosis Regression (CHI-SQUARE) trial evaluated the effect of infusing different doses of CER-001 on 6 occasions following ACS. The trial failed to demonstrate a beneficial effect on the primary imaging end point.¹⁶ However, a subsequent reanalysis in anatomically matched arterial segments demonstrated plaque regression at the lowest dose (3 mg/kg), with a greater effect observed in patients with a larger plaque burden at baseline.¹⁹ Based on these findings, the CER-001 Atherosclerosis Regression Acute Coronary Syndrome Trial (CARAT) was specifically designed to investigate the effect of CER-001 infusions at the 3-mg/kg dose in patients with ACS and a high coronary plaque burden.

Methods

Study Design

The CARAT trial was designed by the South Australian Health and Medical Research Institute in collaboration with the sponsor. The formal trial protocols are available in Supplement 1. Independent ethics boards at each of the 34 participating centers approved the protocol and patients provided written, informed consent. An independent, unblinded data monitoring committee reviewed safety during the study.

The design of the trial has been described previously.²⁰ In brief, eligible patients were 18 years or older, had an ACS event requiring a clinically indicated coronary angiogram and a target vessel deemed suitable for intravascular ultrasonography (IVUS) imaging. This was defined as a major epicardial coronary artery with a maximum lumen stenosis 50% or less, no prior revascularization, and percent atheroma volume (PAV) greater than 30% in its proximal 10 mm on screening by the core laboratory. Patients were excluded if they had uncontrolled diabetes or hypertension; a triglyceride level greater than 500 mg/dL (to convert to millimoles per liter, multiply by 0.0113); or heart failure, renal dysfunction, or liver disease.

The baseline IVUS examination was performed in conjunction with a clinically indicated coronary angiogram at the time of ACS. The details of image acquisition and analysis methods have been summarized previously.^{13,21,22,23,24,25,26,27,28,29} The baseline IVUS examination was evaluated by the Core Laboratory at the University of Adelaide, Adelaide, Australia, to determine whether image quality was acceptable and whether PAV was greater than 30% in the proximal 10 mm of the imaged artery.

Patients who met all clinical, angiographic, and IVUS eligibility criteria and no exclusion criteria were randomized in a 1:1 allocation ratio to treatment with CER-001, 3 mg/kg, or placebo, administered by intravenous infusion. Randomization was stratified by geographic region and occurred no later than 14 days after the index ACS event. Ten infusions of study medication (CER-001 or placebo) were administered at weekly intervals, beginning on the day of randomization. Patients attended clinic visits at the time of each infusion. A follow-up IVUS examination within the same coronary artery was 7 to 21 days after the final infusion.

Analysis of Ultrasonography Images

Analysts at the Core Laboratory, blinded to the treatment status of patients, performed measurements of the lumen and external elastic membrane by manual planimetry on cross-sectional, digitized images, spaced 0.5 mm apart, within a matched segment of artery. We have previously reported the accuracy and reproducibility of this method.^{13,21,22,23,24,25,26,27,28}

Several measures of plaque burden were calculated. The primary efficacy measure, PAV, was calculated as follows:

where EEM_area is the cross-sectional area of the external elastic membrane and Lumen_area is the cross-sectional area of the lumen. The change in PAV was calculated as the PAV at 78 days (within 7-21 days of the final infusion) minus the PAV at baseline. A secondary measure of efficacy, normalized total atheroma volume (TAV), was calculated as follows:

where the mean plaque area in each image was multiplied by the median number of images analyzed in the entire cohort to compensate for differences in segment length between patients. Change in normalized TAV was calculated as the TAV at 78 days minus the TAV at baseline. Change in atheroma volume from baseline to day 78 was also determined in the 10-mm segment that contained the greatest atheroma volume at baseline. Regression was defined as any decrease in PAV or TAV from baseline.

Changes in plaque composition using virtual histology analysis were investigated as an exploratory end point in arteries that were imaged with a 45-MHz rotational catheter (Revolution, Volcano Corporation). This imaging method reconstructs the radiofrequency signal, captured at the peak of the R-wave, to generate a color-coded map that distinguishes necrotic core, dense calcium, fibrofatty, and fibrotic plaque components using echoPlaque 4.3 (Indec Medical Systems), after exclusion of acoustic shadow produced by the catheter. Absolute and percentage plaque measures of each virtual histology parameter were calculated using the trapezoidal rule.³⁰

Efficacy End Points

The primary efficacy end point was the change in PAV from baseline to day 78 (within 7-21 days of the final infusion) as described in previous sections. Secondary efficacy end points included the change in TAV throughout the entire vessel analyzed (normalized to the median vessel length) and within the 10-mm segment containing the greatest atheroma volume from baseline to day 78 and the percentage of participants demonstrating regression of PAV and TAV, defined as any reduction in the parameter from baseline. Exploratory end points included the change in both absolute volume and percentage plaque occupied by individual components determined by virtual histology analysis (necrotic core, dense calcium, fibrofatty, and fibrous).

Statistical Analysis

All statistical analyses were performed using Stata, version 14.2 (StataCorp). A modified intent-to-treat analysis was performed, including all patients who had received at least 1 dose of study drug and evaluable imaging at both points. For normally distributed continuous variables, means and standard deviations are reported. For variables not normally distributed, medians and interquartile ranges are reported. Self-reported race/ethnicity of participants was recorded. The IVUS efficacy parameters are reported as median (interquartile range) and treatment groups compared using analysis of covariance on rank-transformed data with adjustment for baseline value and geographic region. Absolute and percentage changes in laboratory parameters were determined as the difference from baseline to follow-up.

For the change in the primary efficacy parameter, PAV, a sample size of 124 patients in each treatment group provided 86% power at a 2-sided α of .05 to detect a nominal treatment difference of 1.0% assuming a 2.6% SD. Assuming a withdrawal rate of 15%, 146 randomized patients were required. All reported P values are 2-sided. A P value less than .05 was considered statistically significant.

Results

Patient Characteristics

The disposition of patients enrolled in the study is illustrated in the Figure. From August 15, 2015, to November 29, 2016, at 34 centers, 301 patients were randomized to receive study drug, 150 to the placebo group and 151 to the CER-001 treatment group. Two hundred seventy-two patients (90.4%) had evaluable IVUS imaging at both baseline and follow-up. Of these patients, 137 were in the placebo group and 135 were in the CER-001 group. The mean number of infusions administered was 9.7, with 257 (94.5%) receiving all 10 infusions of study drug. Table 1 reports the baseline characteristics of patients with evaluable IVUS data. In association with an ACS index event (non–ST-elevation myocardial infarction, 47% [n = 128]; ST-elevation myocardial infarction, 36% [n = 98]; and unstable angina, 17% [n =46]), patients (mean age, 59.8 years; 80% men [n = 218] and 96% white [n = 261]) had a high prevalence of atherosclerotic risk factors (hypertension, 67% [n = 182]; diabetes, 20% [n = 54]; smoking, 36% [n = 98]; prior myocardial infarction, 10% [n = 27]; and prior coronary revascularization, 14% [n = 38]). Prior to the index ACS event, 77 patients (28.3%) had been treated with a statin. During the study, 258 patients (94.9%) were treated with a statin, of whom 178 (65.4%) received a high-intensity statin.

Figure. Disposition of Patients During the Course of the Trial.

IVUS indicates intravascular ultrasonography.

Table 1. Baseline Characteristics of 272 Patients in the Population With Evaluable IVUS Imaging at Both Points^a.

Parameter	No. (%)
	Placebo (n = 137)	CER-001 (n = 135)
Age, mean (SD), y	59.1 (9.4)	60.6 (9.5)
Male	113 (82.5)	104 (77.0)
White	131 (95.6)	130 (96.3)
BMI, mean (SD)	29.7 (4.7)	28.6 (4.5)
Hypertension	94 (68.6)	87 (64.4)
Previous PCI	26 (19.0)	13 (9.6)
Previous MI	15 (10.9)	12 (8.9)
Smoking	47 (34.3)	52 (38.5)
Diabetes	32 (23.4)	21 (15.6)
Prior statin use	39 (28.5)	38 (28.1)
Post–ACS statin use	130 (94.9)	128 (94.8)
High intensity	84 (61.3)	94 (69.6)
Moderate intensity	46 (33.6)	34 (25.2)
Baseline medications
Antiplatelet therapy	135 (98.5)	131 (97.0)
Antihypertensive therapy	135 (98.5)	131 (97.0)

Abbreviations: ACS, acute coronary syndrome; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); IVUS, intravascular ultrasonography; MI, myocardial infarction; PCI, percutaneous coronary intervention.

^aClinical characteristics and medication use in patients treated with placebo and CER-001, with evaluable IVUS imaging at both points.

Biochemical Measurements

Table 2 summarizes the baseline and posttreatment laboratory values for the 272 patients who underwent follow-up IVUS imaging. At the day 78 visit, median low-density lipoprotein cholesterol (LDL-C) levels were 74 mg/dL in the placebo group and 79 mg/dL in the CER-001 group (to convert to millimoles per liter, multiply by 0.0259) (P = .15). Median posttreatment HDL-C levels were 43 mg/dL in the placebo group and 44 mg/dL in the CER-001 group (P = .66) and median apoA-I levels were 136.0 mg/dL in both groups (P = .87). Accordingly, HDL-C and apoA-I levels did not change with CER-001 infusions as predicted. Triglyceride levels were numerically lower in the CER-001 group (114.0 mg/dL vs 133.0 mg/dL), but this finding failed to meet statistical significance (P = .07). Median high-sensitivity C-reactive protein levels decreased from 5.1 mg/L to 2.3 mg/L in the placebo group and from 5.2 mg/L to 1.6 mg/L in the CER-001 group (to convert to nanomoles per liter, multiply by 9.524) (P = .55). There were small differences between groups in baseline fasting glucose and hemoglobin A_1c levels. There was no significant change in hemoglobin A_1c from baseline in either group. However, the change from baseline differed marginally between groups (0.1%; P = .04). No other differences between treatment groups were observed in biochemical parameters or blood pressure.

Table 2. Baseline and Receiving Treatment Biochemical Measures and Blood Pressure in the Population With Evaluable Intravascular Ultrasonography Imaging at Both Points (N = 272).

Parameter	Prior to First Study Drug Infusion, Median (IQR)a			End of Study Treatment, Median (IQR)b			Absolute Change, Median (IQR)c
	Placebo (n = 137)	CER-001 (n = 135)	P Valued	Placebo (n = 137)	CER-001 (n = 135)	P Valued	Placebo (n = 137)	CER-001 (n = 135)	P Valued
Total cholesterol, mg/dL	150.0 (133.0 to 181.0)	152.0 (131.0 to 186.0)	.55	150.5 (126.0 to 175.5)	148.0 (130.0 to 177.0)	.58	0 (−21.0 to 17.0)	−4.5 (−20.0 to 15.0)	.74
LDL cholesterol, mg/dL	81.0 (65.0 to 99.0)	86.0 (70.0 to 111.0)	.07	74.0 (55.0 to 96.0)	79.0 (63.5 to 98.0)	.15	−6.0 (−24.0 to 10.0)	−5.0 (−19.0 to 6.0)	.75
HDL cholesterol, mg/dL	40.0 (35.0 to 47.0)	40.0 (34.0 to 47.0)	.61	43.0 (36.0 to 51.5)	44.0 (38.0 to 49.0)	.66	2.0 (−2.0 to 8.0)	2.0 (−2.0 to 8.0)	.89
Triglycerides, mg/dL	129.5 (86.5 to 174.5)	113.0 (91.0 to 163.0)	.29	133.0 (93.0 to 192.0)	114.0 (81.0 to 154.0)	.07	5.0 (−32.0 to 38.0)	4.0 (−26.0 to 35.0)	.45
Phospholipid, mg/dL	180.0 (159.0 to 202.0)	180.0 (156.0 to 205.0)	.95	186.5 (160.0 to 213.5)	183.5 (161.0 to 210.0)	.79	6.0 (−14.0 to 27.5)	2.5 (−17.0 to 27.0)	.73
Free cholesterol, mg/dL	40.0 (33.0 to 45.0)	38.0 (33.0 to 47.0)	.86	38.5 (31.0 to 48.0)	37.0 (32.0 to 46.0)	.98	0 (−6.5 to 6.0)	0 (−7.0 to 5.0)	.86
ApoB, mg/dL	76.0 (62.0 to 93.0)	77.5 (65.0 to 90.0)	.37	73.0 (56.0 to 88.0)	72.0 (61.0 to 87.0)	.69	−1.5 (−16.0 to 8.0)	−4.0 (−13.0 to 6.0)	.81
ApoA-I, mg/dL	128.0 (117.0 to 143.0)	126.0 (115.0 to 144.0)	.86	136.0 (120.0 to 155.0)	136.0 (125.0 to 149.0)	.87	9.5 (−5.0 to 23.0)	5.0 (−5.0 to 22.0)	.42
hsCRP, mg/L	5.1 (2.4 to 10.0)	5.2 (1.8 to 11.9)	.71	2.3 (1.0 to 4.4)	1.6 (0.8 to 3.7)	.15	−2.2 (−7.3 to −0.4)	−2.6 (−7.8 to −0.2)	.55
Glucose, mg/dLc	104.0 (95.0 to 119.0)	98.0 (91.0 to 117.0)	.02	107.5 (95.0 to 128.0)	103.0 (96.0 to 116.0)	.18	1.0 (−10.0 to 13.0)	4.5 (−6.0 to 12.0)	.91
HbA1c, %	5.70 (5.40 to 6.30)	5.60 (5.30 to 6.00)	.02	5.70 (5.40 to 6.15)	5.60 (5.40 to 6.00)	.34	−0.10 (−0.30 to 0.10)	0.00 (−0.10 to 0.20)	.04
Systolic BP, mm Hg	129.0 (120.0 to 141.0)	130.0 (118.0 to 144.0)	.67	130.0 (120.0 to 145.0)	132.0 (122.0 to 145.0)	.53	3.0 (−13.0 to 18.0)	2.0 (−13.0 to 16.0)	.97
Diastolic BP, mm Hg	76.0 (69.0 to 82.0)	77.0 (72.0 to 83.0)	.35	76.0 (70.0 to 86.0)	78.0 (72.0 to 85.0)	.53	2.0 (−5.0 to 9.0)	2.0 (−7.0 to 10.0)	.71

Abbreviations: ApoA-I, apolipoprotein A-I; ApoB, apolipoprotein B; BP, blood pressure; HbA_1c, glycated hemoglobin; HDL, high-density lipoprotein; hsCRP, high-sensitivity C-reactive protein; LDL, low-density lipoprotein.

SI conversion factors: To convert cholesterol levels to millimoles per liter, multiply by 0.0259; glucose to millimoles per liter, multiply by 0.0555; glycated hemoglobin to proportion of total hemoglobin, multiply by 0.01; high-sensitive C-reactive protein to nanomoles per liter, multiply by 9.524; triglycerides to millimoles per liter, multiply by 0.0113.

^aBaseline laboratory variables are presented using median and interquartile range and tested using Wilcoxon rank sum test and represent the final value obtained prior to the first infusion of study drug.

^bEnd-of-study measurements are used for receiving treatment values and tested using Wilcoxon rank sum test and represent values obtained at the end-of-study intravascular ultrasonographic examination, or most recent receiving treatment value when that was not available.

^cChanges are presented as median and interquartile range and tested using analysis of covariance on rank-transformed data with adjustment for baseline value and geographic region.

^dP value for between-treatment group comparison.

Primary and Secondary IVUS End Points

Changes in IVUS measures of plaque burden are summarized in Table 3. The primary efficacy measure, PAV, decreased by 0.41% in the placebo group (P = .005 compared with baseline), but did not change in the CER-001 group (−0.09%; P = .67 compared with baseline; between-groups difference, 0.32%; P = .15). The secondary efficacy measure, TAV, decreased by 6.6 mm³ in the placebo group (P < .001 compared with baseline) and by 5.6 mm³ in the CER-001 group (P < .001 compared with baseline; between-groups difference P = .64). Atheroma volume in the 10-mm segment containing the greatest plaque burden at baseline decreased by 3.0 mm³ in the placebo group (P < .001 compared with baseline) and by 3.5 mm³ in the CER-001 group (P < .001 compared with baseline; between-groups difference P = .51). A similar percentage of patients demonstrated regression of PAV (57.7% vs 53.3%; P = .49) and TAV (70.8% vs 67.4%; P = .50) in the placebo and CER-001 treatment groups, respectively. For all prespecified subgroups, there was no significant interaction with treatment assignment on the primary outcome. This included an exploratory analysis of patients stratified to use of statins prior to their index ACS event, which is limited by small patient numbers in each group (eFigure and eTable 1 in Supplement 2).

Table 3. Primary and Secondary Study IVUS End Points^a.

Parameter	Median (IQR)		P Valueb
	Placebo (n = 137)	CER-001 (n = 135)
Baseline
Percent atheroma volume, %	38.4 (32.4 to 43.2)	37.3 (32.8 to 42.4)	.54
Total atheroma volume, mm3	185.7 (125.3 to 235.7)	177.1 (125.7 to 218.1)	.77
Total atheroma volume most diseased 10-mm segment, mm3	78.3 (55.8 to 98.2)	73.6 (57.4 to 96.1)	.73
Follow-up
Percent atheroma volume, %	37.7 (31.6 to 43.3)	37.4 (32.1 to 42.5)	.95
Total atheroma volume, mm3	173.8 (121.3 to 220.2)	172.5 (119.6 to 215.8)	.85
Total atheroma volume most diseased 10-mm segment, mm3	72.9 (52.4 to 95.1)	67.3 (54.0 to 90.9)	.66
Change from baseline
Percent atheroma volume, %	−0.41 (−1.57 to 0.76)	−0.09 (−1.48 to 1.27)	.15
P value for change from baselinec	.005	.67	NA
Total atheroma volume, mm3	−6.6 (−13.7 to 1.7)	−5.6 (−12.4 to 1.7)	.64
P value for change from baselinec	<.001	<.001	NA
Total atheroma volume most diseased 10-mm segment, median (95% CI), mm3	−3.0 (−6.9 to 1.5)	−3.5 (−7.7 to 1.8)	.51
P value for change from baselinec	<.001	<.001	NA
Patients with regression of percent atheroma volume, No. (%)d	79 (57.7)	72 (53.3)	.49
Patients with regression of total atheroma volume, No. (%)c	97 (70.8)	91 (67.4)	.50

Abbreviations: IQR, interquartile range; IVUS, intravascular ultrasonography; NA, not applicable.

^aMeasures of plaque burden at baseline, end of study, and their change in patients treated with placebo or CER-001.

^bP value for between-treatment group comparison, tested using analysis of covariance on rank-transformed data with adjustment for baseline value and geographic region.

^cP value for change from baseline are tested using Wilcoxon matched-pairs signed rank test.

^dDichotomous data are tested using logistic regression model with adjustment for baseline value and geographic region.

Exploratory Analysis of Plaque Composition

An exploratory analysis of plaque composition was performed in the small number of patients (n = 52) who had evaluable virtual histology imaging at both points (Table 4). There were no differences between treatment groups in absolute or percentage plaque volume categorized as fibrous, fibrofatty, calcific, or necrotic at baseline or at follow-up or in the change of any plaque component from baseline to follow-up.

Table 4. Exploratory Analysis of Plaque Composition Parameters.

Parameter	Median (IQR)
	Baseline			Follow-up			Absolute Change
	Placebo (n = 23)	CER-001 (n = 29)	P Valuea	Placebo (n = 23)	CER-001 (n = 29)	P Valuea	Placebo (n = 23)	CER-001 (n = 29)	P Valueb
Fibrous
Absolute, mm3	43.9 (18.7 to 61.9)	34.4 (22.6 to 51.7)	.76	34.8 (24.1 to 55.5)	36.7 (26.6 to 50.7)	.73	−4.2 (−18.1 to 0)	0.95 (−7.4 to 4.2)	.12
Percent plaque	16.4 (11.5 to 22.0)	15.3 (13.5 to 16.7)	.45	14.5 (9.5 to 19.5)	15.4 (13.0 to 18.8)	.61	−1.9 (−4.8 to 1.8)	0.8 (−2.6 to 2.1)	.19
P from baselinec	NA	NA	NA	NA	NA	NA	.04	.87	NA
Fibrofatty
Absolute, mm3	23.2 (16.5 to 52.1)	40.7 (30.3 to 54.7)	.14	20.3 (12.0 to 70.3)	40.5 (21.3 to 65.0)	.31	0.45 (−9.1 to 13.2)	1.2 (−14.3 to 9.7)	.81
Percent plaque	12.1 (8.7 to 18.1)	15.7 (13.8 to 21.9)	.04	11.2 (7.8 to 23.4)	15.5 (10.8 to 24.8)	.25	0.2 (−2.4 to 4.6)	0.9 (−3.7 to 2.9)	.99
P from baselinec	NA	NA	NA	NA	NA	NA	.76	.89	NA
Dense calcium
Absolute, mm3	2.6 (1.1 to 6.9)	2.3 (1.1 to 6.0)	.91	1.6 (0.4 to 5.7)	3.6 (1.7 to 6.0)	.11	−0.3 (−1.25 to 1.9)	0.1 (−2.6 to 1.8)	.66
Percent plaque	1.4 (0.7 to 2.2)	1.2 (0.5 to 2.9)	.95	0.8 (0.2 to 2.6)	1.4 (0.8 to 2.3)	.12	−0.2 (−0.8 to 0.9)	0 (−0.8 to 0.6)	.62
P from baselinec	NA	NA	NA	NA	NA	NA	.78	.99	NA
Necrotic
Absolute, mm3	11.8 (8.9 to 28.8)	14.5 (6.4 to 25.3)	.85	11.5 (6.5 to 25.9)	13.2 (7.0 to 20.9)	.67	−1.5 (−4.3 to 2.9)	−0.8 (−7.5 to 6.5)	.97
Percent plaque	6.0 (4.1 to 9.5)	6.4 (3.1 to 9.3)	.85	4.6 (3.3 to 9.4)	6.6 (3.3 to 7.8)	.60	−0.6 (−2.0 to 1.7)	−0.1 (−3.1 to 2.2)	.93
P from baselinec	NA	NA	NA	NA	NA	NA	.30	.85	NA

Abbreviations: IQR, interquartile range; NA, not applicable.

^aP value for between-treatment group comparison, tested using Wilcoxon rank sum test.

^bP value for between-treatment group comparison, tested using analysis of covariance on rank-transformed data with adjustment for baseline value and geographic region.

^cP values for change from baseline are tested using Wilcoxon matched-pairs signed rank test.

Laboratory and Clinical Adverse Events

eTable 2 in Supplement 2 describes clinical events, clinical adverse events, laboratory abnormalities, and reasons for study discontinuation. Infusions were well tolerated, with no significant difference in any adverse event observed between the treatment groups.

Discussion

Following reports of plaque regression with a HDL mimetic containing apoA-I Milano dimer,¹³ there has been considerable interest in the potential for HDL mimetics to reduce adverse cardiovascular outcomes in patients with clinically manifest atherosclerotic disease. In this study, we evaluated the effect of 10 infusions of CER-001, a bioengineered, negatively charged pre-β HDL mimetic containing recombinant human apoA-I bound to sphingomyelin and dipalmitoyl phosphoglycerol, on coronary atheroma burden in patients following ACS. Despite using a technique that has consistently demonstrated favorable effects of LDL-C lowering on disease progression with long-term treatment,^21,22,28,29 we found no incremental benefit of 10 weekly administrations infusing CER-001 compared with placebo in patients treated with statins.

Despite strong epidemiology evidence demonstrating an inverse association between HDL-C levels and cardiovascular risk,^2,3 contemporary trials of HDL-C–raising agents have failed to produce clinical benefit in statin-treated patients.^{7,8,9,10,11,12} While the potent cholesteryl ester transfer protein inhibitor, anacetrapib, reduced the rate of cardiovascular events in a large outcomes trial,³¹ this benefit appeared to associate with a reduction in atherogenic lipoproteins rather than an increase in HDL-C. In parallel, increasing interest has focused on the potential for agents that affect the functional quality of HDL rather than the quantity of HDL-C. This is supported by initial observations that a benefit of apoA-I Milano infusion was not accompanied by changes in HDL-C,¹³ by genetic data indicating no association of HDL-C with cardiovascular risk,³² and by reports that the composition of HDL is altered and the function of HDL is impaired in patients with established coronary heart disease.³³

CER-001 is a unique recombinant human apoA-I pre-β HDL mimetic distinguished by its negative charge and sphingomyelin content. It is believed that the negative charge, akin to that of native pre-β HDL, in combination with sphingomyelin, prevents fusion of particles, and renal elimination and results in a more rapid and sustained capacity to mobilize lipid. Infusion of CER-001 favorably modified atherosclerosis in mice^14,15 and circulating markers of lipid mobilization in humans.¹⁶ While an earlier serial IVUS study failed to demonstrate a benefit of infusing CER-001 in patients following ACS,¹⁶ a post hoc reanalysis of that trial using anatomically matched arterial segments suggested atheroma regression when CER-001 was infused at the lowest tested dose (3 mg/kg) to patients with a high plaque burden.¹⁹ Plaque burden is associated with lipid content and inflammation. Although these plaque characteristics may portend a higher risk of cardiovascular events, they may also provide a milieu for greater plaque regression with treatment. For these reasons, patients with high plaque burden were selected for this trial; however, the prior findings could not be replicated when subjected to the rigor of prospective, randomized evaluation.

The reason for the lack of benefit of 10 infusions of low-dose CER-001 is uncertain. Patients in this trial received evidence-based therapy for ACS including statin treatment, with most receiving high-intensity statin treatment. It might be argued that plaque regression cannot be accelerated beyond the rate induced by intensive statin treatment. However, a 2016 trial²⁹ demonstrated that evolocumab, a monoclonal antibody to PCSK9, promoted plaque regression on top of statin treatment that was similar in intensity to that in the current study.

Limitations

The optimal dose of CER-001 remains uncertain. While the first serial IVUS study with CER-001 demonstrated a trend toward an inverse dose-response association, thus informing the dose selection for this trial, it is possible that this reasoning was incorrect. Similarly, the optimal dosing frequency for HDL mimetics is unknown. This trial represents, to our knowledge, the largest number of infusions administered to patients with established coronary heart disease yet did not modify plaque burden. The most favorable stage of atherosclerosis for modification by HDL remains unknown, with some reports suggesting a greater likelihood of benefit at the very earliest stages of the disease process.³⁴ It would seem unlikely that intervention at an early stage of preclinical disease with infusion therapy would merit clinical use. Similarly, it is possible that the heightened inflammatory and oxidative state encountered with ACS may have influenced the ability of CER-001 to favorably modify coronary atherosclerosis. Finally, it is possible that cholesterol efflux to CER-001 came primarily from nonvascular sites, with an inconsequential contribution from cholesterol depots in coronary atheroma. CER-001 continues to undergo evaluation in a range of other, nonacute clinical settings.

These findings challenge established concepts regarding cholesterol efflux by HDL and its role in atherosclerosis. Cross-sectional measures of cholesterol efflux capacity associate with cardiovascular risk in observational cohort studies.³⁵ However, to our knowledge, no pharmacologic agent with demonstrable cholesterol efflux capacity has demonstrated favorable effects on either atherosclerosis or cardiovascular events in randomized clinical trials. Other HDL mimetics continue in clinical development with the goal of translating pharmacologic increases in cholesterol efflux to cardiovascular benefit.³⁶ Native HDL possesses a range of non–lipid-transporting properties that may influence the artery wall.³⁷ It is uncertain whether HDL mimetic agents possess these pleiotropic effects of native HDL, and there is a lack of standardized assays for these functions to be assessed in clinical trials. Nonetheless, the lack of a demonstrable effect of CER-001 on atheroma volume in this trial does not necessarily imply a null effect on vascular physiology or cardiovascular risk.

Several other caveats should be noted. First, it is unknown whether these results would have differed had CER-001 been evaluated in a clinical setting other than ACS. Second, this study focused primarily on the effects of CER-001 on plaque burden. While an exploratory analysis of virtual histology imaging demonstrated no effect of CER-001 on these plaque parameters in a small number of patients, it is possible that alternative approaches to characterize plaque phenotype may have yielded different results.

Conclusions

Infusion of low-dose CER-001 for 10 weeks did not result in regression of coronary atherosclerosis in statin-treated patients with ACS and high coronary plaque burden. While CER-001 is unlikely to be of clinical utility in in this setting, its potential clinical utility in other settings will be determined by ongoing trials.

Supplement 1.

Trial Protocol

Supplement 2.

eFigure. Prespecified Subgroup Analysis of Change in Percent Atheroma Volume

eTable 1. Change in Plaque Burden in Treatment Groups in Patients Stratified According to Prior Statin Use

eTable 2. Clinical and Biochemical Adverse Events and Reasons for Discontinuation in the Safety Population