High-density lipoproteins (HDL) are complex endogenous nanoparticles involved in important functions such as reverse cholesterol transport and immunomodulatory activities, ensuring metabolic homeostasis and vascular health. The ability of HDL to interact with a plethora of immune cells and structural cells places it in the centre of numerous disease pathophysiologies. With a better understanding of the properties of synthetic and reconstituted HDL formulations and an increase in preclinical research, several HDL based nanoformulations have been tested in clinical trials.
Traditionally, HDL cholesterol levels have been used as a biomarker to assess cardiovascular risk. However, there is a shift in approach from simply raising HDL cholesterol levels to improving HDL function. This has led to significant efforts to restore HDL quality and enhance its ability to protect against atherosclerosis. HDL particles are highly enriched in apolipoprotein A-I (ApoA-I), accounting for approximately 70% of the total HDL protein [1]. The self-assembly of HDL as definitive nanoparticles is largely attributed to the structural uniqueness of ApoA-I. ApoA-I interacts with endothelial and immune cells to exert a plethora of anti-inflammatory and protective homeostatic functions, making it a natural choice for HDL nanoparticle design.
The intrinsic characteristics of native HDL have inspired therapeutic strategies for synthetic HDL nanoparticles. Their size, natural and specific targeting properties, long circulating time and stability in vivo make them an ideal drug delivery platform with endogenous protective effects. Apart from these significant advantages, one of the most beneficial aspects associated with synthetic HDL nanoparticles is the elimination of extensive biocompatibility validation. This reduces the tolerability issues which are associated with other classes of non-lipoprotein nanoparticles, their clearance being the most concerning question [2].
The administration of ApoA-I isolated from pooled human plasma was attempted in humans owing to favourable data from mouse models of atherosclerosis by overexpressing ApoA-I [3][4][5][6][7][3-7]. In theory, this treatment mimics synthesis and release of nascent ApoA-I to some extent, which would mark the first step of HDL biogenesis. ApoA-I applied to the blood leads to the formation of pre-beta HDL through lipid incorporation via ABCA1, increasing the pre-beta HDL pool, and it later develops into fully mature HDL nanoparticles, analogous to native HDL biogenesis. The study confirmed its tolerability in humans, but functional protective readouts were inconclusive in terms of improved cholesterol efflux capacity, owing to high catabolism of the free protein [8]. This clearly highlights the need for the lipidation of ApoA-I prior to administration.
Another strategy of harnessing cardioprotective benefits of ApoA-I was inspired from the discovery of a mutant variant of ApoA-I in Milan, which was termed as ApoA-I Milano (ApoA-IM) thereafter. As an example of a single nucleotide polymorphism, the variant was characterized by the replacement of arginine by cysteine at position 173, which appears to be associated with better vascular health in carriers [9][10][9,10]. ApoA-IM administration in atherosclerotic mice showed a significant decline in plaque volume and promoted the resolution of coronary artery disease (CAD) [11][12][11,12]. Based on these promising effects, the first randomized controlled clinical trial was designed with the concept of administering full-length recombinant ApoA-IM complexed with POPC, which was called ETC-216 [13]. The results were favourable in terms of reducing plaque volume and CAD regression. A follow-up trial suggested a strong correlation between decreased atheroma volume and compensatory vascular elastic remodeling in a specific, focal manner [14]. Furthermore, an increased understanding of the pharmacokinetic and pharmacodynamic attributes of such interventions were investigated [15]. A similar trial that used the ApoA-IM variant for human trials was the MDCO 216 MILANO PILOT trial and its phase III trial. However, numerous side effects were reported in a dose-dependent manner in association with ETC-216 administration. The studies using POPC with the MDCO-216 formulation (ApoA-IM-POPC) reported a profound increase in ABCA-1 mediated cholesterol efflux in healthy and stable CAD patients [16][17][18][19][16-19]. The primary differences between the MDCO-216 and ETC-216 studies were the lipid composition of the formulations, protein to phospholipids ratio and the number of patients recruited. Later, it was shown that the MDCO-216 formulation was better tolerated by the immune system than the ETC-216 formulation [20]. However, there is no drug based on the ApoA-IM variant, as the trial has been interrupted several times, e.g., due to difficulties in producing large quantities of the full-length protein, increased production costs, change of company owners, etc. Commercially, the strategy of using recombinant full-length proteins has a few drawbacks. The synthesis of large quantities of the protein is expensive, and maintaining the quality can be a challenging aspect. A reduced half-life of about 3 h and rapid renal catabolism is another major challenge that needs to be overcome. The immuno-stimulation problems of ETC-216 were attributed to presence of host antigens, which puts stricter regulation criteria on the concept of full-length, ApoA-I based therapies.
Following up, wild-type ApoA-I was evaluated for its possible therapeutic potential in HDL-mimicking nanodisc formulations. Combining recombinant ApoA-I with two different types of phospholipids, sphingomyelin and dipalmitoyl phosphoglycerol, CER-001 was synthesized and tested in clinical trials funded by Cerenis Pharmaceuticals. CER-001 has shown considerable potential in mouse models of atherosclerosis [21][22][21,22]. In the first double-blind study of multiple intravenous doses of 3 mg/kg, 6 mg/kg, or 12 mg/kg of CER-001 in 507 randomized subjects, the results were not significant in terms of atheroma volume reduction and cumulative coronary stenosis score [23]. A follow-up trial of with serial infusions of CER-001 at a lower dose of 3 mg/kg was designed with 272 patients with acute coronary syndrome, again failing to show a significant improvement in atherosclerotic plaque regression [24]. In 2017, CER-001 was re-studied in a phase I dose optimization clinical trial. The study showed that a single infusion of 45 mg/kg is safe and efficiently mobilizes cholesterol in 32 healthy volunteers [25]. In patients with homozygous familial hypercholesterolemia, prolonged CER-001 infusions demonstrated beneficial effects by significantly reducing mean carotid vessel wall area, suggesting a possible reversal in arterial wall remodeling [26]. Furthermore, in a phase II trial in patients with familial hypoalphalipoproteinemia, CER-001 significantly increased cholesterol mobilization [27]. However, CER-001 did not improve carotid atherosclerosis in patients with genetically very low HDL levels [28]. Overall, CER-001 proved to be well tolerated, but the dose required to potentially see beneficial effects of CER-001 in CAD remains uncertain.
Studies with isolated human ApoA-I were continued by Commonwealth Serum Laboratories (CSL) Behring, testing ApoA-I conjugated with soy phosphatidylcholine in the ERASE trial. Their first formulation, CSL111, was evaluated as short-term infusions in 171 patients testing a lower and a higher dose (40 and 80 mg/kg, respectively). However, the results were not significant in atheroma characterization and coronary angiography results when compared to the placebo group [29]. Moreover, the treatments were associated with a significant increase in bilirubin in the higher dose group, and hypotension was seen in the lower dose group. Following up on the formulation, CSL112 was first tested in rabbits [30] and was later administered in single and multiple doses in healthy subjects, showing a dramatic and immediate increase in ApoA-I levels, cholesterol efflux capacity and cholesterol esterification [31]. In a randomized, multicentre, double-blind phase IIa study in patients with stable atherosclerosis, a single infusion of CSL112 in groups of three doses, 1.7 g, 3.4 g, or 6.8 g, was well tolerated, and the treatment resulted in significant cholesterol mobilization [32][33][32,33]. The first part of the AEGIS I study conducted in approximately 1200 CAD patients who had recently experienced an acute myocardial infarction event, was designed to characterize the dosing and safety profile of CSL112 [32]. CSL112 at a dose of 6 g dramatically increased cholesterol efflux to about 300%, which was much more compared to all other previous studies [34]. The efflux capacity was reported to be similar in healthy and stable atherosclerotic patients [35] that remained unaffected in conditions of moderate underlying renal complications [36]. This outcome is of specific relevance, as kidney disease severely limits HDL functionality [37][38][37,38]. In the CSL112-related studies, significant emphasis has been placed on improving study design, gaining mechanistic insights, and continuously monitoring off-target/additive drug effects and co-treatment options [39][40][41][39-41]. The CSL112-based studies are one of the few that seek to determine and understand native HDL remodeling and pre-beta HDL formation with exogenous sHDL nanotherapy. The potential beneficial effects of CSL112 are being investigated in an ongoing multicentre, double-blind, randomized, placebo-controlled phase III study (AEGIS II study) including about 17,400 participants [42]. The primary objective of AEGIS II is to investigate the efficacy of CSL112 in reducing the risk of death, myocardial infarction, or stroke from the time of randomization to 90 days in patients with acute coronary symptoms.
Despite the protective and anti-inflammatory effects of full-length ApoA-I, there are disadvantages associated with them, such as high production costs and quality control. The first mimetic peptide developed to study the ApoA-I nanodisc formation, 18A, gave deeper understanding of the HDL nanodisc assembly and paved the way for introducing nuances in the structural design to enhance vascular protective features [43][44][43,44].
The sHDL nanodisc formulation ETC-642 was developed by Esperion and has been extensively tested for its anti-atherogenic characteristics [45][46][47][45-47]. The ETC-642 complex consisted of ESP-2418, a 22 amino acid ApoA-I mimetic peptide complexed to two phospholipids, sphingomyelin and DPPC. The trials in human subjects proceeded to phase I with good tolerability and evidence of LCAT activation [47]. However, the study was discontinued in 2006 due to Pfizer ownership. Following this first mimetic peptide leading to a clinical trial [48], several studies reported anti-inflammatory properties of ETC-642 in mice and in vitro [47][49][50][51][47,49,50,51]. Currently, ETC-642 is not being tested in clinical trials.
The 4F peptide developed by Novartis was tested, for the first time, in oral formulations [52][53][54][55][56][52-56]. The 4F peptide was formulated with both D- and L-amino acids, which had similar biological characteristics but different bio-availabilities due to different helicity and amphipathicity. Here, D-4F was well tolerated in a phase I trial, and preliminary data suggested a potent anti-inflammatory activity [57][58][57,58]. D-4F showed promising results and the study indicated the intestine as the main site of action with plasma concentrations of D-4F being independent of the route of administration. However, L-4F did not show significant positive results in CAD prevention [59]. In vitro studies have demonstrated 4F-mediated HDL remodeling [60]. It should be emphasized here that the peptides were not lipidated before administration.
The ApoA-I mimetic peptide 5A produced by KlineMed. Inc. is currently another promising candidate for atherosclerosis prevention, as shown by in vitro and in vivo data from mouse and primate models [61][62][63][61-63]. 5A peptide incorporated into nanodiscs with sphingomyelin has demonstrated high cholesterol efflux capacity, anti-inflammatory and anti-oxidant properties in the animal models and is under safety profile and pharmacokinetic evaluation in a phase I clinical trial. Table 1 summarizes the current strategies and clinical trials with ApoA-I based HDL nanoparticles and mimetic peptides.
An overwhelming number of studies highlight the role of HDL in cardiovascular disease. Continuous optimization efforts are required to improve the pharmacological characteristics, such as the half-life and targeting accuracy of synthetic HDL nanoparticles, as demonstrated by recent reports on modified and loaded HDL nanodiscs. The focus of new HDL therapeutics has shifted from increasing circulating HDL-C concentrations to improving HDL function, particularly HDL cholesterol efflux capacity, which is impaired in patients with cardiovascular disease. In summary, there is now sufficient evidence from preclinical and clinical studies that the enhancement/restoration of cholesterol efflux by HDL mimetic nanoparticles is a valid strategy to stabilize atherosclerotic plaque through various anti-inflammatory/immunoregulatory processes. In addition, the immunomodulatory effects of HDL nanoparticles on monocytes and macrophages enable multiple cardioprotective properties that improve vascular health. Data from the ongoing phase III AEGIS-II trial testing CSL112 may represent a breakthrough in the treatment paradigm for acute myocardial infarction, particularly in the immediate weeks and months after the event, when patients are at high risk for the recurrence of ischemic events.
Table 1. Summary of HDL nanoparticles and ApoA-I mimetics in clinical trials.
Summary of HDl nanoparticles and ApoA-I mimetics in clinical trials.
HDL |
Protein Type |
Peptide Sequence |
Phospholipid |
Nanoparticle Size |
Ratio |
Administration |
Patients |
Clinical Trial |
Major Findings |
Reason to Discontinue Clinical Trial |
ETC-216 |
Recombinant ApoA-IM |
ApoA-I |
DPPC |
7–30 nm |
1:2.7 (w/w) |
5 weekly, |
57 |
Milano, Phase II |
Significantly decreased atheroma volume |
Change of company ownership; protein production and quality constrain |
MDCO-206 |
Recombinant ApoA-IM |
ApoA-I |
POPC |
7–30 nm |
1:1.1 (w/w) |
5 weekly, intravenous, |
126 |
MILANO-PILOT, Phase III, |
80.4% increase in cholesterol efflux, no increment in plaque regression |
Change of company ownership; protein production and quality constrain |
CER-001 |
Recombinant ApoA-I |
ApoA-I |
SM, DPPG |
7–13 nm |
1:2.7 (w/w) |
10 weekly, intravenous, 3 mg/kg with statins and 6 weekly, intravenous, |
507 |
CHI SQUARE, Phase II NCT01201837 and CARAT, Phase II NCT2484378 |
Failed to promote coronary atherosclerosis regression |
Insufficient beneficial effects in CAD |
CSL111 |
Plasma purified human ApoA-I |
ApoA-I |
Soy PC |
7–30 nm |
1:150 molar ratio |
4 weekly, |
111 |
ERASE Phase II, NCT00225719 |
Significant improvement in the plaque characterization index, elevation in liver enzymes in 80 mg/kg group |
Modified and improved to CSL112 |
CSL112 |
Plasma purified human ApoA-I |
ApoA-I |
Soy PC |
7–13 nm |
1:55 molar ratio |
4 weekly, intravenous, 2 g and 6 g |
1258 |
AEGIS-I, Phase II, NCT02108262 |
No renal/hepatic toxicity, well tolerated |
_ |
CSL112 |
Plasma purified human ApoA-I |
ApoA-I |
Soy PC |
7–13 nm |
1:55 molar ratio |
4 weekly intravenous, 6 g |
17400 |
AEGIS-II, Phase III, NCT03473223 |
Results yet to be concluded |
_ |
ETC-642 |
ApoA-I mimetic- 22 amino acids, single helix |
P-V-L-D-L-F-R-E-L-L-NE-L-L-E-AL-K-Q-K-L-K |
DPPC |
7–13 nm |
1:1:1 molar ratio |
4 weekly intravenous, |
28 |
Phase I |
Safe and well tolerated, LCAT activation |
Change of company ownership |
D-4F |
ApoA-I mimetic peptide, 18 amino acids, bihelical |
Ac-D-W-F-K-A-F-Y-D-KV-A-E-K-F-K-E-A-FNH 2 |
Non- |
- |
- |
Three separate groups, sequential, ascending doses 100 mg, 300 mg, or 500 mg, oral |
104 |
Phase I, NCT00907998 |
Increased hydrophobicity, blocks absorption of production of oxidized lipids, anti-inflammatory |
Insufficient |
L-4F |
ApoA-I mimetic peptide, 18 amino acids long, bihelical |
4F enantiomer, |
Non- |
- |
- |
Single and 7-daily infusions, Intravenous, subcutaneous |
176 |
Phase I, NCT00568594 |
Similar to 4F but shorter circulating time due to proteolysis |
Insufficient |
5A |
ApoA-I mimetic peptide, 37 amino acids, bihelical |
D-W-L-K-A-F-Y-D-K-V-A-E-K-L-K-E-A-F-P-D-W-A-K-A-A-Y-D-K-A-A-E-K-A-K-E-A-A |
SM |
- |
1:8 |
5 weekly, intravenous, |
64, currently recruiting |
Phase I, NCT04216342 |
ABCA-1 specificity, results yet to be concluded |
_ |
Legends: dipalmitoyl phosphatidylcholine (DPPC), 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC), dipalmitoyl phosphatidylglycerol (DPPG), palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-linoleoyl phosphatidylcholine (PLPC), phosphatidylcholine (PC), sphingomyelin (SM).
References: