Lipoprotein(a) in Atherosclerotic Diseases: From Pathophysiology to Diagnosis and Treatment

Lipoprotein(a) (Lp(a)) is a low-density lipoprotein (LDL) cholesterol-like particle bound to apolipoprotein(a). Increased Lp(a) levels are an independent, heritable causal risk factor for atherosclerotic cardiovascular disease (ASCVD) as they are largely determined by variations in the Lp(a) gene (LPA) locus encoding apo(a). Lp(a) is the preferential lipoprotein carrier for oxidized phospholipids (OxPL), and its role adversely affects vascular inflammation, atherosclerotic lesions, endothelial function and thrombogenicity, which pathophysiologically leads to cardiovascular (CV) events. Despite this crucial role of Lp(a), its measurement lacks a globally unified method, and, between different laboratories, results need standardization. Standard antilipidemic therapies, such as statins, fibrates and ezetimibe, have a mediocre effect on Lp(a) levels, although it is not yet clear whether such treatments can affect CV events and prognosis. This narrative review aims to summarize knowledge regarding the mechanisms mediating the effect of Lp(a) on inflammation, atherosclerosis and thrombosis and discuss current diagnostic and therapeutic potentials.


Introduction
Atherosclerosis constitutes a chronic arterial disease, initiated very early in life, while its progression appears slow, with its clinical manifestations peaking during the fifth and sixth decades of life [1,2]. Atherosclerosis is the underlying cause of most cardiovascular diseases (CVD), which account for more than 17.9 million deaths per year [3]. Inflammation has a pivotal role in this context as both innate and adaptive immunoinflammatory mechanisms are involved in the atherosclerotic properties of the arterial wall [4]. Atherosclerotic cardiovascular diseases (ASCVD) are highly dependent on modified lifestyle factors, such as obesity, type 2 diabetes mellitus (T2DM) and metabolic syndrome [5]. Moreover, older age, hypertension, dyslipidemias, cigarette-smoking and physical inactivity are some of the traditional risk factors leading to acceleration of atherogenic processes and early cardiovascular (CV) events [5].
However, genetic predisposition is a key factor in ASCVD risk and progression, and, despite effective management of atherosclerotic risk factors, residual atherosclerotic CV risk appears to be significant. Elevated lipoprotein(a) (Lp(a)) in plasma is an independent, mainly genetically determined causal risk factor for CVD [6]. Lp(a) was first described in 1963 by Kåre Berg as a unique lipoprotein similar to low-density lipoprotein (LDL) [7]. Lp(a)     Moreover, several studies proclaim the detrimental effect of smaller apo(a) isoforms [44]. Smaller apo(a) particles, resulting in smaller Lp(a) molecules, appear to (1) increase the capacity of bound oxidized phospholipids; (2) accumulate more in blood vessel walls through increased lysine-binding ability and interaction with fibrin; (3) increase inhibition of plasmin activity, resulting in greater thrombogenicity [45]. Furthermore, smaller apo(a) isoforms show synergistic action with small-dense LDL and oxidized LDL (OxLDL) particles [46].

LPA Gene Variations and Atherosclerotic Disease
The LPA gene, which encodes the apo(a) component of the Lp(a) particle and is located on the long arm of chromosome 6 within 6q2. 6-2.7 [47], shows homology up to 70% with the plasminogen gene [29]. The heterogeneity in apo(a), and, consequently, Lp(a) size, is based on the number of KIV2 copies, which play a central role in controlling the circulatory levels of Lp(a) [48,49]. Fewer KIV2 repeats reduce the apo(a) size, leading to higher Lp(a) levels as hepatocytes can produce smaller apo(a) particles at higher rates [50].
The apo(a) isoforms are modified by many SNPs distributed over the entire range of allele frequencies, with very strong effects on Lp(a) concentrations [51]. Genome-wide association studies identify two SNPs in the LPA gene, which are very strongly and independently correlated with increased Lp(a) levels and higher risk for CVD [44,52-54]: Rs3798220 and rs10455872 SNPs. However, the mechanisms underlining the effect of SNPs on apo(a) size and Lp(a) levels are still vaguely understood. Large studies reveal that LPA gene polymorphisms, such as rs783147, rs3798220 and rs10455872, are also strongly associated with CV lesions, such as increased carotid intima-media thickness and impaired endothelial function, suggesting a direct role of LPA gene SNPs in early atherosclerotic changes [55].

Measurement of Plasma Lp(a) Concentration and Standardization
While Lp(a) has started to emerge as a risk factor for cardiovascular disease, with recommendations for measuring it at least once during lifetime, especially in high-risk populations [56,57], issues regarding standardization and validity of methods used to measure it are still present. Since the size of Lp(a) is highly variable and largely determined by the apo(a) size, which varies across different populations, there is inter-individual and intra-individual variability as most individuals are carriers of two different apo(a) alleles [40,58,59]. Accordingly, kits that aim to quantify the mass concentration of Lp(a) are more prone to variation compared to molarity-measuring tests, leading to diverging values of different mass-targeting kits, even within the same population with a standard Lp(a) molar concentration [60,61]. Despite this issue and the apo(a)-insensitive quantifying methods, [61,62] commercial kits measuring Lp(a) in mg/dL instead of estimating its molarity in nmol/L are still amply encountered in practice and the literature [63,64]. Moreover, converting the mass concentration of Lp(a) to its molar equivalent (mg/dL to nmol/L) cannot produce accurate results, although attempts have been made and a rough 2-2.5× conversion factor has been proposed [65].

Attachment of Lp(a) to Arterial Wall
Lp(a) exerts its atherogenic actions when transferred from circulation to the arterial wall. In general, the arterial influx of lipoproteins depends on arterial wall permeability, the concentration of lipoprotein in plasma and arterial blood pressure [66]. Lp(a) has been detected in human vessels and is concentrated mainly extracellularly in the intima and subintima [8]. Anchoring of Lp(a) depends on its two components: its lipoproteinic structure and the lysine binding sites of apo(a) [67,68]. Notably, Lp(a) is accumulated in the arterial wall to a greater extent than LDL, as derived from the relative amounts of apo(a) and apoB-100 that have been detected in early atherosclerotic plaques [69,70]. It seems that Lp(a) remains extracellularly due to its interactions with molecules of the extracellular Molecules 2023, 28, 969 5 of 30 matrix (ECM) and other cells [71]. The lysine binding sites of apo(a) contribute to Lp(a) hooking to ECM [72]. Indeed, transgenic mice with a mutation in KIV10 lysine binding sites display decreased amounts of Lp(a) within vessel walls [73]. A two-part interaction between Lp(a) and proteoglycan decorin has been described, comprising an electrostatic bond between apoB-100 and glycosaminoglycan (GAG) chain of decorin and a hydrophobic one between proteoglycan's protein core and apo(a). The second interaction may explain the differences in arterial wall affinity distinguishing Lp(a) from LDL [74]. Detection of decorin in atherosclerotic arterial wall argues for a possible role of this interaction in atherogenesis [75]. Furthermore, possible ligands of Lp(a) are alpha-defensins. These proteins are neutrophil-derived, found in human atherosclerotic plaques and cluster with Lp(a). These stable complexes do not traverse the cytoplasmic membrane of the endothelial cells, and this is another mechanism that maintains Lp(a) extracellularly [76]. Nevertheless, Lp(a) is also integrated into cells, mainly in macrophages in order to form foam-cells. A multitude of interactions favor this formation. At first, there is interplay with fibronectin, a glycoprotein of the ECM. Lp(a) linked to fibronectin alone or in combination with heparin can enter macrophages and enhance lipid-driven atherogenesis [77][78][79]. In addition to this, the cholesterol content of macrophages augments internalization and degradation of Lp(a) and apo(a), and this calcium-dependent process is independent of LDL, scavenger, LRP or plasminogen receptors and is not driven by proteoglycans of the cell membrane [80]. A different mechanism demands contribution of ECM, enzymes, macrophages and vascular smooth muscle cells (VSMCs). Macrophages release sphingomyelinase that drives, in collaboration with lipoprotein lipase, adhesion of LDL and Lp(a) to bovine aortic smooth muscle cells (SMCs) and ECM. Interestingly, proteoglycan chondroitin ABC lyase attenuates this effect. SMCs do not integrate the aggregates of LDL and Lp(a), in contrast with mouse macrophages that are converted to foam cells [81]. Finally, oxidative modification of Lp(a) by malondialdehyde, in a similar way to LDL, promotes integration and degradation of this lipoprotein, mediated by the scavenger receptor of human monocyte-macrophages [82].

Effect of Lp(a) on Expression of Adhesion Molecules, Chemotactic Factors and Other Cytokines
Lp(a) favors initiation of atherogenesis by modulating recruitment of inflammatory cells in the vessel wall. Expression of adhesion molecules, vascular cell adhesion molecule 1 (VCAM-1) and E-selectin [83] in cultured human coronary endothelial cells [84] and intercellular adhesion molecule 1 (ICAM-1) in cultured human umbilical vein endothelial cells [85] is upregulated by Lp(a). The last effect of ICAM-1 could be partially attributed to inhibition of transforming growth factor-β (TGF-β) by Lp(a) [85]. Furthermore, Lp(a), along with β2-integrin macrophage-1 antigen (Mac-1), facilitate attachment and infiltration of monocytes. In order to achieve this, Lp(a) activates transcription factor NFkappaB, confirming inflammatory involvement in atherogenesis [86]. Another contribution of Lp(a) in orchestration of atherogenesis is the effect on chemotaxis. A study on human monocytes showed that Lp(a) exhibits chemotactic properties. The incapability of LDL to induce such responses and the inhibitory effect of plasminogen and inactivated plasmin on the Lp(a)-driven chemotaxis in monocytes suggest that the lysine binding sites of apo(a) might be responsible. The same study proposes a cGMP-dependent mechanism for this effect [87]. Lp(a) accelerates chemotaxis also indirectly by driving human endothelial cells to secrete monocyte chemotactic protein (MCP) [88]. Moreover, Lp(a) and especially the C-terminal region of apo(a) induce release of interleukin-8 (IL-8) by human THP-1 macrophages [89], promoting infiltration of neutrophils [90]. More precisely, this effect is mediated by oxidized phospholipids (OxPLs) bound to apo(a), and their interaction with apo(a) requires the lysine-binding site in KIV10 [18,91]. Except for IL-8, Lp(a) also increases expression of interleukin-1β (IL-1β) [92] and tumor necrosis factor-α (TNF-α) by macrophages, multiplying inflammation of the arterial wall [42,87,88].

The Impact of Oxidation in Lp(a) Function
Even though oxidized phospholipids (OxPLs) are mainly produced by oxidation of LDL, OxPLs are closely associated with Lp(a) in human circulation [93]. Low Lp(a) levels act beneficially as they remove OxPLs from the plasma [94] and induce their cleavage via a lipoprotein-associated phospholipase A2 (Lp-PLA2) mechanism [95,96]. In contrast, high Lp(a) levels result in an excessive amount of OxPLs in the arterial wall [94]. Thus, certain atheromatic actions of Lp(a) are carried out via OxPLs and include chemotaxis, formation of foam cells, enhancement of inflammation and plaque instability [93,94,97,98].

Effect on Vascular Smooth Muscle Cells
Lp(a) promotes proliferation of human-cultured VSMCs by blocking conversion of plasminogen to plasmin. This blockage does not enable plasmin-mediated activation of TGF-β, which is an autocrine inhibitor of SMC growth [99,100]. Additionally, apo(a) provokes a concentration-dependent chemorepulsion to SMC that is dependent on RhoA and integrin αVβ3 but unrelated to TGF-β [101].
Moreover, emerging evidence highlights the role of extracellular vesicles in pathogenesis of atherosclerosis and atherothrombosis [102][103][104]. Platelet-derived exosomes have a pivotal role in hemostasis and thrombosis because of their interaction with several vascular cell types (i.e., endothelial cells, vascular smooth muscle cells) and intercellular communication [105,106]. Finally, a single study shows that high levels of Lp(a) partially mediate SMCs and valvular interstitial cells calcification via inducing release of calcifying extracellular vesicles [107].

Lp(a) and Plaque Vulnerability
Lp(a) also affects the stability of atherosclerotic plaques. Metalloproteinases and elastases are enzymes, detected in atherosclerotic sites, that are responsible for the split of Lp(a) into two fragments, F1 and F2 [108,109]. F2 fragment, but not F1, is the one that interacts with fibrinogen, fibronectin and decorin, which are key molecules involved in the actions of Lp(a) [74,109]. For instance, matrix metalloproteinase-12 (MMP-12) is related to production of F1 fragment from injected human Lp(a)/apo(a) [108]. In parallel, IL-8, whose release is favored by Lp(a), downregulates expression of inhibitors of metalloproteinases [110]. Thus, a balance between Lp(a) and metalloproteinases must be achieved to avoid both the inflammatory effects of Lp(a) and the degradation of ECM by MMPs. Interestingly, Lp(a) itself also modifies plaque stability [101]. Lp(a) augments expression of micro-PAR and ICAM-1, receptors for vitronectin and fibrinogen, respectively, and thus enhances monocyte linkage to ECM. Lp(a) also upregulates expression of urokinase and urokinase receptors on monocytes and, therefore, activation of plasmin that promotes ECM shrinkage [111]. A different mechanism refers to OxPLs, as a part of apo(a), that facilitate apoptosis of endoplasmic-reticulum-stressed macrophages and, therefore, plaque necrosis [112,113]. Lp(a), thus, possesses a crucial role not only in the early phases of atherosclerosis but also later during the sequence of events that result in destabilization of atherosclerotic plaques ( Figure 3). thus enhances monocyte linkage to ECM. Lp(a) also upregulates expression of urokinase and urokinase receptors on monocytes and, therefore, activation of plasmin that promotes ECM shrinkage [111]. A different mechanism refers to OxPLs, as a part of apo(a), that facilitate apoptosis of endoplasmic-reticulum-stressed macrophages and, therefore, plaque necrosis [112,113]. Lp(a), thus, possesses a crucial role not only in the early phases of atherosclerosis but also later during the sequence of events that result in destabilization of atherosclerotic plaques ( Figure 3).

Oxidized Phospholipids and Lp(a)
The reassociation between atherosclerosis and inflammation has already been documented [4]. Lp(a), as a source of free fatty acids and monoacylglycerols (via cleavage by lipoprotein lipase), can be detected in the immune system and provoke local inflammation [114]. Nevertheless, oxidative modulation of lipoproteins, including Lp(a), is a crucial mediator that can determine the evolution of inflammation. For instance, OxPLs present on OxLDLs can initiate sterile inflammation [115]. As with LDL, Lp(a), as the main carrier of OxPLs, modulates inflammation. Low Lp(a) concentrations remove OxPLs from plasma, yielding a probably beneficial effect as OxPLs can be considered as damageassociated molecular patterns (DAMPs), and, thus, the immune system is activated to restrain them [116]. Macrophage scavenger receptors gather OxPLs via Lp(a) as an antibody against apo(a) is efficient to abolish this interaction, with the lysine binding site in KIV10 being again the key mediator. Moreover, the activity of Lp(a) as platelet-activating factor acetylhydrolase (PAF-AH) [95], recently known as lipoprotein-associated phospholipase A2 (Lp-PLA2), modifies the concentration of OxPLs as they convert them to oxidized fatty acids and lysophosphatidylcholine (lysoPC), reducing their concentration. Yet, OxPLs, bound to apo(a), compete for the catalytic site of Lp-PLA2, and, as a result, attenuate its enzymatic activity and increase OxPL levels in an autocrine loop. The absence of apo(a) was found to diminish this effect, highlighting the apo(a) contribution to this effect [96]. Therefore, both beneficial and detrimental effects of Lp(a) appear to exist, related to regulation of OxPLs bioavailability in a dose-dependent manner: while an anti-inflammatory role is suggested for lower levels, this association is reversed when higher Lp(a) levels are reached [94].

Effect of Lp(a) in Chemotaxis
Concerning chemotaxis, Lp(a), beyond its effect on IL-8, interacts also with monocyte chemoattractant protein-1 (MCP-1), both directly through binding of OxPL with MCP-1 and indirectly by inducing its production [88,117]. Remarkably, a cGMP-driven chemotactic property of Lp(a) has been documented independent of MCP-1 and other chemokines [87]. Additionally, production of I-309, a chemotactic factor that attracts monocytes, is induced by Lp(a) in human umbilical vein endothelial cells [118]. On the other hand, experiments in apo(a) transgenic mice showed that apo(a) inhibits neutrophil recruitment due to reduced levels of neutrophil chemokines, macrophage-inflammatory protein-2, CXCL1 and CXCL2 [119,120]. Thus, Lp(a) interacts with different mediators of chemotaxis, promoting inflammation.

Effect of Lp(a) in Monocyte Phenotype
In vitro studies demonstrate that monocytes in the presence of Lp(a) skew towards a pro-inflammatory phenotype, exhibiting more adhesive and migratory properties. This effect was found to relate with the apo(a), especially with the site that binds OxPLs [42]. More precisely, apo(a) facilitates adhesion of monocytes on type I collagen and stimulates matrix metalloproteinase-9 (MMP-9) production and thus collagen degradation and macrophage infiltration [121]. These pro-inflammatory, M1-type macrophages, under the influence of Lp(a), release CXCL10 (IP10) chemokine in order to activate T-helper-1 (Th1) lymphocytes and natural killer (NK) cells [122]. Lp(a) also induces secretion of cytokines by macrophages, such as IL-1β, TNF-α and IL-6, multiplying, in this way, the inflammatory effect [42,123].

Lp(a) Induced Changes in the Inflammatory Milieu
Lp(a) was found to induce transcription factor nuclear factor-kappaB (NF-κB) in monocytes interacting with β2-integrin Mac-1, leading to augmented infiltration of monocytes [86]. Another study suggests that Lp(a) exerts this effect via autotaxin activity. Autotaxin converts lysophosphatidylcholine, a component of OxPLs, into lysophosphatidic acid, and the latter is responsible for activation of NF-κB cascade [124]. Inversely, some inflammatory cytokines can affect Lp(a) gene expression by interacting with inflammatory response elements in LPA gene. For instance, Il-6 increases, whereas TGF-β and TNF-α diminish, the gene expression in primary monkey hepatocyte cultures [125]. Therefore, there is a dynamic relationship between Lp(a) and inflammatory processes (Table 1). Cross-sectional 267 patients with CAD.
• Lp(a) levels did not show any differences among subgroups of patients (one-vessel disease, two-vessel disease and three-or-more-vessel disease (multi-vessel).

•
The percentage of apo(a) isoforms of low molecular weight (655 kDa) and the percentage of subjects with at least one apo(a) isoform of low molecular weight were significantly correlated with increasing number of coronary vessels stenosed. • Lp(a) plasma levels were significantly associated with early atherogenesis in subjects with LDL cholesterol levels above the population median (3.3 mmol/L). • Apo(a) phenotype did not differ in subjects with and without early carotid atherosclerosis. • Apo(a) phenotype was found a strong risk predictor of advanced atherogenesis (carotid stenosis >40%), especially LMW apo(a) phenotypes when associated with high Lp(a) plasma concentrations.   • Serum Lp(a), oxidized LDL antibody (oLAB) and plasma total TFPI in patients with CAD were found to be significantly higher than in the control group. • A significant positive correlation between serum Lp(a) and plasma total TFPI levels in CAD was found.

Lp(a) Modifies the Properties of Endothelial Cells
Lp(a) can affect the transcriptional profile of endothelial cells and skew them towards a more inflammatory phenotype. Lp(a) induces expression of adhesion molecules, such as VCAM-1, E-selectin and ICAM-1, on cultured human endothelial cells [84,85]. Inhibition of TGF-beta by Lp(a) could explain in part the effect on ICAM-1 expression [85]. Notably, Lp(a) was found to accelerate senescence of endothelial cells and formation of reactive oxygen species (ROS) in human aortic endothelial cells, in combination with upregulation in expression of p53 and p21, molecules that regulate cell-cycle [149]. Moreover, ROS produced by endothelial cells in the presence of Lp(a) disrupts the integrity of the endothelial cell barrier by increasing the permeability of the endothelial cell monolayer [150]. Similarly, the apo(a) component of Lp(a) leads, via the Rho/Rho-kinase-dependent signaling pathway, to increased phosphorylation of myosin light chains and rearrangement of actin cytoskeleton, resulting in enhanced contraction of endothelial cells and loss of cell contact [151,152]. Lp(a) additionally impairs activity of endothelial progenitor cells (EPCs). Apo(a) was found to attenuate adhesion and migration properties of EPCs [153]. EPCs coexistence with Lp(a) led to a decrease in CD31-positive capillaries and did not improve ischemic limb perfusion. In vitro, generation of capillary tubes was halted by full-length recombinant apo(a) and the urinary fragments of apo(a) [154]. Apo(a) both impairs angiogenesis signaling pathways and also provokes endothelial cell apoptosis, along with modifying nuclear factors in endothelial cells via OxPLs [32]. For instance, Lp(a) activates a focal adhesion kinasemitogen-activated protein kinase (MAPK)-dependent pathway, enhancing in that way migration and proliferation of endothelial cells [155]. Lp(a) also suppresses TGF-β, which has been found to inhibit endothelial cell migration in vitro [155,156]. Therefore, Lp(a) affects endothelial cell proliferation and permeability and also exerts anti-angiogenetic properties, interfering with different mediators.

Effect of Lp(a) on Endothelial-Dependent Vascular Tone
At first, Lp(a), via AKT-mediated nuclear translocation of β-catenin, leads to cyclooxygenase 2 expression in endothelial cells and release of prostaglandin E2 (PGE2), a key molecule that modulates vasodilation [157]. On the other hand, Lp(a), especially the oxi-dized form, reduces dose-dependently the expression of inducible nitric oxide synthase (iNOS), both at mRNA and protein levels, in lipopolysaccharide/interferon-stimulated mouse macrophages [158]. In the same regard, oxLp(a) particles, as well as oxLDL ones, were found to inhibit Ach-induced endothelium-dependent vasodilation [159]. The inhibitory effect of oxLp(a) on vasorelaxation can be multiplied by glycation. Indeed, the ox-Lp(a) and the glycated Lp(a) suppressive effects are prevented by administration of factors that act as O2 scavengers, suggesting that oxygen is responsible for the inactivation NO [160]. High Lp(a) levels affect negatively endothelium-dependent vasodilation, even in the absence of atherosclerotic lesions, suggesting a direct Lp(a) inhibitory effect or an effect mediated by increased endothelial permeability that enables direct access of acetylcholine to the VSMCs [161]. Interestingly, relatively high Lp(a) plasma levels were related to upregulated endothelium-dependent vasoconstriction in the presence of L-NMMA, a NO synthase inhibitor. A possible mechanism can be compensatory increased basal production and secretion of NO to Lp(a)-induced atherogenesis [162].

Role of Lp(a) on Thrombogenicity
The partial similarity of apo(a) with plasminogen is not enough to explain the significant role of Lp(a) in the pathways regulating hemostasis. Lp(a) affects discrete key points in primary and secondary hemostasis as well as in fibrinolysis.

Activation and Aggregation of PLTs
The intervention of Lp(a) in the function of platelets has already been documented since 1985 as Lp(a), via apo(a), attenuated collagen-and ADP-driven platelet aggregation and platelet production of thromboxane A2 when co-incubated with human platelets [163]. Lp(a) does not interact with platelet LDL binding sites, so it is Lp(a) that inhibits collagen-or ADP-mediated platelet aggregation [164]. There is an epitope, unique to apo(a) in Lp(a), the single arginyl-glycyl-aspartyl (RGD) epitope, which mainly mediates the suspensory effect of Lp(a) [165]. At first, Lp(a) anchors to resting platelets at a site different from the RGD epitope on the IIb subunit of the IIb/IIIa receptor [166] and then binds via the RGD epitope of apo(a) to the RGD binding site on the IIb protein of the GPIIb/IIIa receptor of agoniststimulated platelets [165]. GPIIb/IIa is the receptor for fibrinogen, but RGD-type ligands such as apo(a) can interact with the receptor even though fibrinogen is already attached and can displace fibrinogen in an allosteric way [167]. The result is reduced collagen-and ADP-stimulated platelet aggregation [168]. Furthermore, in resting platelets, a role for cyclic-AMP has been described as low concentrations of Lp(a) in vitro induce elevation of intracellular c-AMP above basal levels and restrict collagen-stimulated platelet aggregation, whereas, at high concentrations, levels of c-AMP return to basal even though platelet aggregation continues to decline. This effect is also apo(a)-mediated [169]. Nevertheless, the role of Lp(a) and its components in activation and aggregation of PLTs has been a matter of dispute over the years. A study showed that lysine binding sites and not GPIIb/IIIabound fibrin affect binding of recombinant-apo(a) to platelets [170], while a later study presented that the lysine binding sites of Lp(a) are not correlated to the anti-aggregatory effect of Lp(a) in vitro [171]. The first study also exhibited a noteworthy finding: intact Lp(a) or recombinant apo(a) facilitate aggregation of platelets to subaggregant doses of arachidonic acid, whereas recombinant apo(a) does not affect the response to low doses of collagen or thrombin [170]. In line with this, native Lp(a) does not influence collagen or thrombin or ADP-driven platelet activation [172,173]. In contrast, Lp(a) was found to augment thrombin-receptor-activating hexapeptide (SFLLRN, TRAP)-induced platelet activation [173]. A different approach for the role of Lp(a) may derive from its activity as platelet-activating factor acetylhydrolase (PAF-AH) [95]. Lp(a) inhibits PAF-induced platelet activation but not via PAF-AH-action of Lp(a), and the crucial mechanism involved remains interaction with GPIIb/IIa and fibrinogen [174].

Interraction of Lp(a) with the Tissue Factor (TF)
Administration of Lp(a) or recombinant apo(a) (r-apo(a)) to monocytes upregulates production of TF and its presence in the cell membrane. The involvement of both the integrin αMβ2 and NFκB signaling pathways has been suggested as possible mechanisms [175].
Tissue factor pathway inhibitor is derived mainly from the endothelium and has been colocalized with vascular SMC of human atherosclerotic plaques along with apo(a) [176][177][178]. Lp(a) and apo(a) can bind to recombinant TFPI (rTFPI) in vitro in a lysine-dependent way and Lp(a) is able to impair rTFPI activity and endothelial cell surface TFPI activity in vitro independent of plasminogen [177].

Inhibition of Fibrinolysis by Plasmin
Apo(a) and fibrin co-localize in the arterial wall, predisposing an interaction between the two [179]. Lp(a) has an antifibrinolytic effect that may be attributed to multiple mechanisms restricting the bioavailability and action of plasmin, an enzyme that degrades fibrin. The precursor of plasmin is plasminogen, and it is significant to mention that there are two types of plasminogen: Glu-plasminogen and Lys-plasminogen, derivatives of Glu-plasminogen via plasmin-cleavage [180]. Homology with plasminogen is suggested as the most prominent linkage of Lp(a) with fibrinolysis. Lp(a) and plasminogen share similar lysine-binding sites on the fibrinogen/fibrin molecule. The size of apo(a) isoform determines inversely the affinity of Lp(a) to fibrin, with smaller size leading to greater affinity [181]. Therefore, the relative concentration of Lp(a) particles with small apo(a) isoforms can demonstrate a marker for the risk concomitant of high levels of Lp(a) [182]. Notably, plasmin catalyzes binding of Lp(a) with fibrinogen and fibrin. Therefore, a fibrin thrombus during its formation can bind with Lp(a) and simultaneously activate plasminogen, with produced plasmin further enhancing Lp(a) binding. Conversely, Lp(a) is bound to the fibrin surface, inhibiting binding of Glu-plasminogen and Lys-plasminogen to the same position and thus attenuating degradation of fibrin [183]. Moreover, reduced adherence of plasminogen to endothelial cells [184,185] and platelets can be attributed at least in part to Lp(a) [186]. Lp(a) impairs linkage of plasminogen to annexin, a plasminogen receptor of platelets and endothelial cells, and, therefore, prevents the activation of plasminogen in the surface of these cells [88]. Additionally, Lp(a) impairs the binding of t-PA to the platelet surface [174]. Moreover, factor XIIIa can catalyze linkage of Lp(a) and fibrinogen, and, remarkably, this coagulating factor co-localizes with Lp(a) in human atherosclerotic plaques [187]. This emphasizes that the interaction of Lp(a) with fibrinogen is not only dependent on the two of them but also on other factors [187]. Another and equally important mechanism described involves impairment of plasminogen activation. The effect prevails t-PA-induced activation of Glu-plasminogen and affects less t-PA activation of Lys-plasminogen and urokinase-mediated activation [188]. The initial formation of a binary complex between tPA and fibrin facilitates formation of a ternary complex between substrate (plasminogen), enzyme (tPA) and cofactor (fibrin) that precedes cleavage of plasminogen by tPA and release of plasmin [189,190]. Of note, apo(a) can regulate formation of the ternary complex as it can induce generation of a quaternary complex that includes the former. The quaternary complex is characterized by a reduced turnover number, suggesting a possible mechanism for the inhibitory effect of Lp(a) in plasminogen activation [191]. An alternative explanation may be that Lp(a) has direct crossover with t-PA [192] or Lp(a), impairing formation of the binary complex as it competes with tPA as a candidate for fibrin linkage [193]. There are also contradictory data demonstrating that Lp(a) does not interact with t-PA [194] or that apo(a) enhances instead of diminishes the bound of plasminogen with fibrin and tPA-induced plasminogen activation [195]. Finally, experiments in mice confirm the requirement of Lp(a) for t-PA-induced fibrinolysis [196]. Except for t-PA, streptokinase-induced plasminogen activation [197] is impaired, as well as urokinase, showing the broad effect of Lp(a) on fibrinolysis [188].

Increased Expression of Plasminogen Activator Inhibitor
Lp(a) also affects activation of plasminogen indirectly by inducing production of inhibitory molecules by surrounding cells. Expression of plasminogen activator inhibitor-1 (PAI-1) is upregulated in endothelial cells [198] in the presence of Lp(a), while monocytes from male patients with Lp(a) hyperlipidemia exhibit increased PAI-2 mRNA and protein [199].

Lp(a) and Neointimal Hyperplasia
Neointimal hyperplasia is implicated in vascular restenosis. Evidence shows that Lp(a) is correlated with vascular restenosis. Increased levels of Lp(a) were found to predict vein graft stenosis after bypass procedure [200], and apo(a) moiety has been detected in diseased vein grafts, enhancing a possible role of Lp(a) [9]. Furthermore, several studies have demonstrated an association between circulating Lp(a) and restenosis after percutaneous transluminal coronary angioplasty (PTCA). Lp(a) has been described as an independent predictor of restenosis, with higher levels correlated with higher risk of restenosis [113,[201][202][203] or with an increased degree of restenosis [204]. Nevertheless, studies that reveal no significant correlation between Lp(a) and restenosis after PTCA have been developed so far [205,206].
Lp(a) may be involved in different steps of this pathophysiological mechanism. At first, Lp(a), as mentioned above, anchors to arterial wall and exhibits inflammatory properties [70]. Lp(a), in coordination with β2-integrin Mac-1, facilitates infiltration of monocytes [207], and, of note, Mac-1 blockade was found to attenuate experimental neointimal thickening [208], revealing the inflammatory character of neointimal growth. Lp(a) can also be associated with macrophages that incorporate into the injury site [209], and, indeed, it has been documented that neointimal tissue is lipid-laden [210,211], and areas of restenosis 5 years post stent implantation included cholesterol clefts, necrotizing foam cells and inflammatory cells [212]. In addition, Lp(a) increases infiltration of leukocytes by increasing expression of VCAM-1 [83].
A further step to restenosis is neointimal hyperplasia, either promoted by thrombus or by Lp(a) itself [213]. Neointimal hyperplasia is a physiologic healing response to internal or external injury to the blood vessel wall, involving all three arterial layers and characterized by proliferation of SMCs [214]. Thrombus, including platelets, mediates release or prolongs exposure to chemotactic or growth factors, such as platelet-derived growth factor, and also favors synthesis of ECM, thus preparing the local milieu and creating a "scaffold" for intimal hyperplasia [201,213,215]. Furthermore, coagulating factors such as thrombin or Xa have been proven to drive mitosis for SMCs in vitro [216,217]; thus, Lp(a) indirectly enhances neointimal hyperplasia by promoting thrombosis. Interestingly, TFPI, which, as already indicated, can be inactivated by Lp(a), may affect not only thrombosis but also SMC proliferation [177]. TFPI has been found to attenuate intimal hyperplasia post angioplasty, and this was mainly attributed to inhibition of thrombosis [218][219][220].
It is also noteworthy to mention that Lp(a) exerts direct effects on SMC proliferation. Lp(a) inhibits plasminogen activation, plasmin generation and, subsequently, production of active TGF-β, which impairs the migration and proliferation of SMCs [100,221]. Furthermore, when a neutralizing antibody silenced the effect of TGF-β, Lp(a), but not apo(a), induced SMC proliferation, suggesting that the LDL-particle of Lp(a) may exert also a mitogenic role [222]. Several Lp(a)-positive lesions did not contain thrombus, highlighting the significant effect of Lp(a) on neointimal growth.

Predictive Value of Lp(a) Levels
Circulating plasma Lp(a) levels increase very soon after birth, on the seventh postnatal day, and reach a constant concentration even in a few months of life [223]. Individual Lp(a) concentrations are relatively stable throughout the lifetime and range widely from <1 to >200 mg/dL in the general population [224]. Many studies suggest that women are more prone to increased Lp(a) concentrations [225,226]. According to a Danish general population cohort study (Copenhagen General Population Study), an estimated approximately 20% of the population have high concentrations corresponding to >42 mg/dL, which have long been linked to increased risk of atherosclerotic diseases [227]. Worth noting as well are the differences in Lp(a) plasma levels between different populations as Lp(a) seems to be lowest in Caucasian patients and highest in patients of African ethnicity [228].
Because >90% of circulating Lp(a) levels are genetically determined, little effect from diet and environment is detected, and concentrations over a lifespan do not vary considerably [229,230]. According to "2016 ESC/EAS Guidelines for the Management of Dyslipidaemias", Lp(a) concentrations should be considered for Lp(a) screening in selected high-risk cases for reclassification of subjects with borderline risk, with a class IIa indication and a level of evidence: C [231], specifically subjects with a family history of premature CVD, familial hypercholesterolemia and a family history of premature CVD and/or elevated Lp(a) levels, when Lp(a) is above the 80th percentile (50 mg/dL) [6]. Moreover, patients with recurrent CVD events despite treatment for lipid-lowering and, finally, subjects with ≥5% 10-year risk of fatal CVD according to SCORE [232] should also be treated as high-risk patients, and Lp(a) levels should be examined [231]. The US National Lipid Association indicates the positive predictive power of Lp(a) measurement and provided similar recommendations on screening methods, adding in the screening procedures for patients with 10-19% Framingham risk according to the 2012 Canadian Cardiovascular Society recommendations [233,234]. Additionally, HEART UK (Hyperlipidaemia Education and Atherosclerosis Research Trust UK), in the recent consensus statement, suggests that individuals with calcific aortic valve stenosis should be evaluated for their Lp(a) levels, having employed data from the large ongoing Copenhagen General Population Study [52,235] ( Table 2).

Lp(a) Should be Measured Once Over a Lifespan in Individuals with:
(1) premature CVD

Lp(a)-Lowering Treatment
Converging evidence proclaims the role of Lp(a) in cardiovascular disease, including myocardial infarction, ischemic stroke and calcific aortic valve disease [237,238]. Its action seems to be mediated not only by its lipid-carrying content but mainly by its ability to carry and deliver oxidized phospholipids (OxPL) directly to tissue targets, which are converted to lysophosphatidic acid by autotaxin, infiltrate the endothelium and promote inflammation [238,239]. Data show the Lp(a)-lowering effect of new antilipidemic agents and establish Lp(a) as a potential new target for decreasing cardiovascular risk as lifestyle modifications are unlikely to have any effects on Lp(a) levels because of the primary genetic basis [240][241][242].
According to recent evidence, conflicting data exist regarding the effect of statins on Lp(a). Although statins remain one the most effective and safest drug category for primary prevention of ASCVD, a recent study revealed a mean 11% increase in Lp(a) levels with their use [63, 243,244]. Moreover, the ILLUMINATE trial revealed that, in high-risk CVD patients, Lp(a) levels are positively and dose-dependently correlated with atorvastatin dosage [245]. Most meta-analyses on the impact of different types and dose schemas of statins show no clinically significant reduction in Lp(a) levels [246]. Despite this effect of statins, the most recent "European Atherosclerosis Society consensus statement" suggests that statin therapy should not be discontinued as their cardioprotective action overcomes any risk associated with increased Lp(a) plasma concentrations [6].
Data on the role of ezetimibe on Lp(a) circulating levels are not solid. A recent metaanalysis of seven randomized controlled trials shows an Lp(a) reduction by 7%, although this reduction is considered unable to reduce the Lp(a)-related risk of CVD events [247]. However, another large meta-analysis, collecting more robust data from 10 randomized placebo-controlled clinical trials, demonstrated that ezetimibe therapy had no effect on altering plasma Lp(a) concentrations, either as a monotherapy or in combination with a statin [248].
Niacin (nicotinic acid) has been used for over 50 years [249] for reduction in CVD events and mortality, being one of the most effective available therapies for raising HDL [250]. Niacin is also the only currently approved treatment for Lp(a) reduction, acting by silencing apo(a) gene expression in hepatocytes [247]. This effect is dose-dependent and leads to a 25% to 38% reduction in Lp(a) levels when niacin is administered at a 2 to 4 g daily dosage, respectively. However, niacin has not yet shown any effect on CVD reduction [251]. Although a large meta-analysis of 14 randomized placebo-controlled clinical trials reported a significant reduction by 23% in plasma Lp(a) concentration, the prognostic relevance of this effect has yet to be clarified [252], while the Lp(a)-lowering effect of niacin has not been linked to any clinical benefit, in terms of ASCVD events, so far [253].
Proprotein convertase subtilisin/kexin type 9 inhibitors (PCSK9i) upregulate LDL receptor (LDLR) activity [254]. Plasma PCSK9 levels are associated with Lp(a) particles in humans and mice [255]. Otherwise, in animal studies, circulating PSCK9 levels are positively associated with apo-B synthesis [256,257]. Recently, PCSK9i achieved a clinically meaningful reduction in serum Lp(a) concentrations as, in a meta-analysis of 41 studies (n = 64,107 randomized patients), Lp(a) levels were reduced by 26.7% [258]. The Lp(a) reduction may be attributed to reduced availability of lipoproteins containing apo-B to link with Lp(a), uptake and clearance of Lp(a) by LDLR or other hepatic receptors under a state of low LDL-C levels [257,259]. The reduction in Lp(a) under PSCK9i treatment may also be caused by reduced apo-B and Lp(a) synthesis [259]. The role of apo-B availability for Lp(a) synthesis is also supported by studies, with antisense oligonucleotides against apo-B synthesis showing parallel Lp(a) reduction [257,260]. Noteworthily, PCSK9i have proven their efficacy by decreasing incidence of acute coronary syndromes and CV deaths in patients with CAD [261], while, in experimental studies, their interaction with several proinflammatory factors is crucial [262]. Specifically, Alirocumab reduced the risk of major adverse cardiovascular events by 0.6% for each 1 mg/dL reduction in Lp(a) levels independent of LDL cholesterol reduction [261]. In the FOURIER trial, patients with high baseline Lp(a) levels experienced a greater drop in Lp(a) levels (up to 27%) after receiving evolocumab [224]. Several clinical trials have been conducted to investigate the role of PCSK9i therapy on Lp(a)-lowering [263][264][265][266][267][268]. Ge et al., in a large meta-analysis of seven clinical trials, exhibited an average reduction in Lp(a) levels of~18%, with a follow-up period ranging from 8 to 78 weeks [269].
Recently, antisense oligonucleotides (ASO) designed to inhibit apo(a) mRNA seem promising, with preliminary results pointing to an even more substantial reduction reaching up to 90% [270]. Subcutaneously injected ASO are taken up by hepatocytes, where they bind to the apo(a)-mRNA, causing its breakdown and thus inhibiting apo(a) synthesis [271]. ISIS-APO(a)Rx, a second-generation antisense drug designed to reduce synthesis of apo(a) in liver, also shows promising results in Lp(a) reduction [272]. ISIS-APO(a)Rx reduces plasma Lp(a) in a dose-dependent manner, along with the associated OxPL. Further clinical trials are required to prove the efficacy of ISIS-APO(a)Rx to reduce CV events and calcific aortic valve stenosis [272].

Conclusions
Lp(a) is recognized as a risk factor for atherosclerotic and non-atherosclerotic cardiovascular disease, with its levels being largely genetically determined and mediated by variations in the LPA gene locus. Oxidized phospholipids carried by Lp(a) adversely affect various pathways (i.e., vascular inflammation, endothelial function and thrombogenicity), contributing to atherosclerosis progression. Despite the recognized role of Lp(a) in atherosclerosis, its measurement lacks a globally unified method, hampering any effort to appropriately identify individuals at higher risk. Moreover, until now, only scarce evidence exists to support the clinical benefit of Lp(a) level reduction for the few available agents. Fortunately, knowledge of the pathophysiologic mechanisms of Lp(a) synthesis and action has fueled research interest and oriented drug manufacturing efforts to more prominent approaches in management of Lp(a)-related CV risk, with novel therapeutic options anticipated in the coming years.