Targeting Lipoprotein(a): A New Era in Cardiovascular Disease Prevention

Genetic Determinants

Between 70% and > 90% of circulating Lp(a) levels are genetically determined.⁶ Located on chromosome 6, the LPA gene, which encodes apolipoprotein(a) [apo(a)], is the principal gene recognized to regulate plasma Lp(a) concentrations. Two codominantly expressed LPA alleles determine Lp(a) levels.^6,7 LPA is believed to have evolved from a duplication of the plasminogen gene (PLG), which explains the remarkable structural similarities between the two proteins.^8,9 However, unlike PLG, which encodes five kringle subtypes, LPA encodes only kringle subtypes IV and V, both structurally similar to those of PLG, as well as an inactive protease domain.^8,10

Genes outside the LPA locus also influence Lp(a) levels. Variants of the APOE ɛ2 allele and CETP are associated with lower Lp(a) concentrations, whereas a novel variant in the APOH gene that encodes β2-glycoprotein I, which has been linked to PCSK9 biology, has been associated with higher Lp(a) levels.^11,12,13

These genes, however, explain only part of the variability in Lp(a). The genetic architecture of Lp(a) exhibits substantial heterogeneity, arising from ancestral differences in the number of apo(a) kringle IV type 2 repeat (KIV-2) alleles, as well as from single-nucleotide polymorphisms in coding and noncoding regions. Despite this complexity, accurate polygenic risk scores have been developed that can explain up to ~60% of trait variance.¹⁴ However, most of these scores have been derived from European populations and often do not transfer well to other ancestral groups.^14,15

Molecular Structure

Lp(a) is an LDL-like particle consisting of a lipid core composed primarily of cholesteryl esters, with smaller amounts of triglycerides, surrounded by an outer shell containing apolipoprotein B-100 (apoB-100) covalently bound to apo(a).^16,17 The apo(a) moiety is characterized by KIV and KV domains. The KIV domain comprises ten distinct subtypes (KIV-1 through KIV-10) but is unique in that the KIV-2 subtype occurs in a variable number of repeats.⁸ The number of KIV-2 repeats determines the size of apo(a) isoforms and contributes to the marked heterogeneity observed in Lp(a) particles: It was noted in the UK Biobank that KIV-2 length explained nearly 60% of variance in Lp(a) levels.^17,18 Smaller isoforms typically contain fewer KIV-2 repeats (≤ 22), whereas larger isoforms contain more repeats. Of note, there is an inverse relationship between the number of KIV-2 repeats and circulating Lp(a) concentrations plausibly related to the efficiency of secretion of apo(a).^17,19

Pathophysiologic Mechanisms

Lp(a) is synthesized primarily in the liver, as the LPA gene is not expressed in other organs^8,20 The constituent components of Lp(a) contribute to its proinflammatory,²¹ prothrombotic,¹⁷ and proatherogenic²¹ properties. The apo(a) moiety is structurally homologous to plasminogen but lacks protease activity. As a result, some studies have suggested that apo(a) may interfere with fibrinolysis, contributing to antifibrinolytic effects.⁸ In addition, the apo(a) component may interact with platelet protease-activated receptor-1, potentially promoting atherothrombosis.²²

Lp(a) also contributes to inflammation and atherogenesis through its enrichment with oxidized phospholipids (OxPL), making it the primary carrier of OxPL among apoB-containing lipoproteins.^21,23 This is mediated by binding sites for OxPL within the apo(a) moiety.¹⁹ These properties promote delivery of Lp(a)-associated OxPL to sites of vascular injury, including the arterial endothelium and aortic valve leaflets, leading to upregulation of cell adhesion molecules and recruitment of inflammatory cells, including macrophages, and ultimately culminating in vascular inflammation, lipid accumulation, and calcific remodeling.¹⁹ The presence of the apo(a) moiety and coexisting OxPL is a key distinction from LDL that makes Lp(a) more atherogenic; Lp(a) confers a five- to sixfold higher estimated risk of coronary heart disease than an equimolar LDL level.^24,25,26

Observations that inhibition of interleukin-6 signaling with ziltivekimab is associated with dose-dependent reductions in Lp(a) levels of up to 25%²⁷ support a potential Lp(a)–inflammation–atherogenesis axis. Beyond its role in plaque formation, Lp(a) has also been implicated in the development of vulnerable plaque phenotypes, including noncalcified necrotic-core plaques with thin fibrous caps, which are more prone to rupture.²⁸ Observational studies support this: Lp(a) levels are associated with progression of low-attenuation plaque on coronary computed tomography angiography.²⁹ This may reflect the preferential association of Lp(a) with noncalcified plaque.³⁰ It has been shown that the cardiovascular risk associated with Lp(a) may be independent of coronary artery calcium burden.^31,32

Epidemiology

Elevated Lp(a) levels are estimated to affect approximately one in five individuals globally, corresponding to more than one billion people worldwide.^15,33 Lp(a) levels tend to approach adult levels around age five, although some reports suggest they may increase until early adulthood and then generally remain stable.³⁴ Changes in Lp(a) levels occur after menopause and in the setting of inflammatory disease states.^33,35,36 Substantial variation exists across racial and ethnic groups, with increasing median levels and prevalence of elevated Lp(a) observed in the following order: East Asians, Europeans, Southeast Asians, Latin Americans, Middle Eastern populations, South Asians, and Black individuals. In the INTERHEART Study, median Lp(a) levels among patients with coronary artery disease were 9.8 mg/dL in Chinese, 11.5 mg/dL in Latin Americans, 12.9 mg/dL in Southeast Asians, 14.7 mg/dL in Europeans, 18.1 mg/dL in Arabs, 18.9 mg/dL in South Asians, and 27.1 mg/dL in Africans. Corresponding median levels among controls were 7.8 mg/dL, 9.6 mg/dL, 10.2 mg/dL, 11.5 mg/dL, 13.8 mg/dL, 15.3 mg/dL, and 27.2 mg/dL, respectively.³⁷ These racial differences in Lp(a) levels become evident from a young age.³⁸

The differences are thought to arise partly from variation in the inverse relationship between apo(a) isoform size and circulating Lp(a) levels, with the strongest association observed in individuals of African and Latin American ancestry.³⁷ Differences in the dominance patterns of alleles in the LPA locus may also contribute; smaller apo(a) alleles appear to be more frequently dominant in individuals of European ancestry compared with those of African ancestry.³⁹ Furthermore, protein-altering intronic variants identified in the UK Biobank explain a substantial proportion of the difference in median Lp(a) levels between participants of European and African ancestry, with Lp(a)-lowering variants present in approximately 43% of European alleles compared with only 13% of African alleles.¹⁸ Interestingly, significant heterogeneity in Lp(a) levels exists even within racial and ethnic groups,^40,41 highlighting the need for further investigation.

Association with Cardiovascular Disease

Lp(a) is an established causal risk factor for coronary artery disease, peripheral artery disease (PAD), heart failure, aortic valve stenosis (AS), aortic aneurysm, and ischemic stroke.³³ Following his discovery of Lp(a) in 1963, Kåre Berg reported an association between elevated Lp(a) levels and coronary heart disease in 1974.⁴² This finding was subsequently corroborated by numerous observational studies.^43,44 Since then, higher Lp(a) has been associated with AS, PAD, stroke, and other diseases.^45,46

In 2009, Børge G. Nordestgaard and colleagues demonstrated a potentially causal association between elevated Lp(a) levels and myocardial infarction using Mendelian randomization.⁴⁷ The associations between Lp(a) and other CVDs have now been reported in many large genome-wide association studies and Mendelian randomization studies.⁴⁸ The strongest association between Lp(a) and ASCVD is observed with coronary artery disease, whereas associations with PAD and stroke are less robust. For example, higher stroke risk with elevated Lp(a) levels was observed only in women in the Atherosclerosis Risk in Communities Study⁴⁹ but only in men in the Cardiovascular Health Study.⁵⁰ Similarly, large observational studies in PAD cohorts have not consistently demonstrated higher incidence or worse outcomes with elevated Lp(a) levels.^51,52 Ongoing randomized clinical trials evaluating Lp(a)-lowering therapies may further clarify the role of Lp(a) in ASCVD.

Beyond ASCVD, elevated Lp(a) levels are also potentially causally associated with calcific AS. This association is thought to result from the delivery of OxPL by Lp(a) to injured aortic valve leaflets, which promotes osteogenic differentiation of valvular interstitial cells and accelerates valvular calcification.⁵³ Observational associations between elevated Lp(a) levels and AS risk have been supported by genome-wide association studies and Mendelian randomization analyses.^54,55 In addition to predicting incident AS, elevated Lp(a) levels are associated with more rapid disease progression and a greater likelihood of valve replacement, independent of baseline disease severity.⁵⁶ In a secondary analysis of the FOURIER trial, higher baseline Lp(a) levels were associated with increased risk of AS-related events, and reductions in Lp(a) were associated with lower risk.⁵⁷

Assay Standardization

Measurement of Lp(a) can be technically challenging.⁵⁸ Key issues include assay standardization, reporting units, and the choice of biological sample. These challenges with standardization largely arise from the structural complexity of the molecule. In particular, variability in the number of KIV-2 repeats in apo(a) introduces substantial size heterogeneity, which complicates immunoassay standardization.⁵⁹ The use of appropriate calibrators and isoform-insensitive assays is therefore critical to minimize measurement bias related to apo(a) size variability.⁶⁰ Current recommendation is the use of an assay traceable to official reference standards.⁶¹ Efforts are ongoing to standardize and harmonize these measurements.⁶² Another important challenge relates to reporting units. Laboratories commonly report Lp(a) concentrations in either mg/dL or nmol/L. Current recommendations favor reporting in nmol/L because this unit reflects particle concentration and accounts for differences in apo(a) isoform size, whereas mg/dL reports mass concentration.⁶³ Although there have been efforts to convert levels directly between mg/dL and nmol/L, this is discouraged, especially since measurements from both identify high-risk individuals.⁶³

Finally, it remains uncertain whether the source of the blood sample (ie, plasma versus serum) significantly influences Lp(a) measurement. In a recent analysis of paired samples from 100 ARIC participants, we observed strong concordance between plasma and serum Lp(a) measurements.⁶⁴ However, additional studies on this are needed.

Screening Recommendations

Guidelines on Lp(a) testing have evolved with growing evidence of the cardiovascular risk associated with elevated Lp(a) levels and the development of targeted Lp(a)-lowering therapies. The European Society of Cardiology/European Atherosclerosis Society,⁶⁵ Canadian Cardiovascular Society,⁶⁶ National Lipid Association (NLA),³⁴ and American College of Cardiology/American Heart Association (ACC/AHA)⁶³ lipid guidelines all recommend at least one-time measurement of Lp(a) in adults as part of ASCVD risk assessment. The NLA recommends selective screening in high-risk children with familial hypercholesterolemia, first-degree relations with premature ASCVD, ischemic stroke of unknown cause, or first-degree relatives with elevated Lp(a).³⁴ These recommendations reflect the strong causal association between elevated Lp(a) levels and CVD. However, despite evidence of cost-effectiveness of Lp(a) testing,⁶⁷ testing rates are estimated at only 1% to 2%,⁶⁸ but they are increasing.

Intraindividual measurements can vary by as much as 20%,⁶⁹ but Lp(a) levels tend to remain relatively stable over time and are largely resistant to lifestyle modification; therefore, a single measurement is generally sufficient for most individuals. However, levels may increase with the transition to menopause or in the presence of secondary conditions such as kidney disease, liver disease, thyroid dysfunction, pregnancy, or certain medications.⁶³ In such cases, repeat measurement may be warranted, particularly if initial values were borderline. More than 50% of individuals with borderline Lp(a) have high levels on repeat testing, whereas individuals with low or high Lp(a) levels tend to remain in their respective categories..⁷⁰ Cascade testing is recommended in families with elevated Lp(a), especially when this coincides with a personal or family history of premature ASCVD, AS, or familial hypercholesterolemia.⁶³

Risk Thresholds and Stratifications

The risk of ASCVD associated with elevated Lp(a) increases in a continuous manner.⁴⁸ Compared with levels < 75 nmol/L, levels of 75 nmol/L to 125 nmol/L are associated with “borderline-high” risk (approximately 1.2-fold higher ASCVD risk), whereas levels > 125 nmol/L confer “high” risk (approximately 1.4-fold higher). The thresholds for borderline and high risk in mg/dL are 30 mg/dL and 50 mg/dL, respectively. More pronounced elevations confer substantially greater risk. A twofold, threefold, and fourfold increase in ASCVD risk has been observed at Lp(a) concentrations ≥ 250 nmol/L, ≥ 350 nmol/L, and ≥ 430 nmol/L, respectively.⁴⁸ Lp(a) levels ≥ 125 nmol/L (or ≥ 50 mg/dL) are therefore recognized by the NLA and American Society of Preventive Cardiology as a cardiovascular risk enhancer when personalizing lipid-lowering therapy, and more recently by the ACC/AHA for individuals with an estimated 10-year ASCVD risk of 3% to 5% using the PREVENT-AHA Risk Calculator (Predicting Risk of cardiovascular disease EVENTS).^34,63

Current Therapies

Although no pharmacologic therapies are specifically approved for lowering Lp(a), several agents approved for other indications have been shown to reduce Lp(a) levels (Table 1). Current management of elevated Lp(a) focuses primarily on optimizing modifiable ASCVD risk factors, including those outlined in the AHA’s Life’s Essential 8 framework. Adherence to healthy lifestyle factors has been associated with up to a 67% reduction in ASCVD risk among individuals with elevated Lp(a).⁶³ Additionally, although statins do not meaningfully lower Lp(a) concentrations, analyses from the JUPITER (Justification for the Use of Statins in Primary Prevention: an Intervention Trial Evaluating Rosuvastatin) study demonstrated that participants with elevated Lp(a) treated with high-intensity statin had approximately 30% to 40% reductions in cardiovascular event risk, underscoring the importance of intensive LDL-C lowering.⁶³ Current guidelines recommend intensive management of CVD risk factors.^34,63

Among currently available lipid-lowering therapies, niacin and proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors have demonstrated Lp(a)-lowering effects. Niacin reduces Lp(a) levels by up to approximately 37%,⁷¹ likely through downregulation of transcriptional activity at the LPA promoter.⁷² However, this reduction did not translate into improved cardiovascular outcomes in the HPS2-THRIVE (Heart Protection Study 2–Treatment of HDL to Reduce the Incidence of Vascular Events) trial despite a significant absolute reduction in Lp(a) levels (-9.5 mg/dL; P = .006).⁷³

PCSK9 inhibitors reduce Lp(a) levels by approximately one quarter.⁷⁴ The underlying mechanism remains incompletely understood, although it is thought to involve enhanced clearance and possibly reduced production of Lp(a).⁷¹

Evolocumab, alirocumab, and inclisiran—respectively studied in the LAPLACE–TIMI 57 (Lipid-Lowering Arm of the Assessment of PCSK9 Inhibition in the Thrombolysis in Myocardial Infarction 57) and FOURIER (Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk) trials,^75,76 the ODYSSEY Outcomes trial,⁷⁷ and the ORION trial—have each demonstrated sustained reductions in Lp(a).

Although both evolocumab and alirocumab have demonstrated reductions in cardiovascular events, it is uncertain whether these benefits are independent of their LDL-C-lowering effects. A post hoc analysis of the FOURIER trial and prespecified analysis of the ODYSSEY Outcomes trial suggest that the benefits of these medications could be partly from their Lp(a)-lowering effects.^76,77 For patients with elevated Lp(a) and LDL-C levels above target, or at high risk of ASCVD, PCSK9 inhibitors should be considered as additional therapy given their ability to lower both LDL-C and Lp(a) levels.^63,65,66

Lipoprotein apheresis represents another therapeutic approach for lowering Lp(a). This therapy removes apoB–containing lipoproteins, including Lp(a) and LDL, from the circulation and has been reported to reduce Lp(a) levels by approximately 30% to 68%.⁷⁸ Several observational studies have suggested that lipoprotein apheresis may reduce cardiovascular events in patients with markedly elevated Lp(a), although randomized clinical trial evidence remains limited.^33,78 Lipoprotein apheresis is approved by the United States Food and Drug Administration for patients with Lp(a) > 60 mg/dL (or 130 nmol/L) and LDL-C > 100 mg/dL in the presence of familial hypercholesterolemia and coronary artery disease or PAD.^34,63

Finally, post hoc analyses of several primary prevention trials suggest that aspirin therapy may reduce cardiovascular events in individuals with elevated Lp(a). Although these reports are plausible given the prothrombotic properties of Lp(a), prospective trials specifically evaluating aspirin therapy in this population are needed.^34,63

Emerging Lipoprotein(A)-Lowering Therapies

Novel therapies acting directly on Lp(a) have shown great potency in lowering Lp(a) levels. Agents in phase 3 development include antisense oligonucleotides (ASOs, such as pelacarsen), small interfering RNA (siRNA) therapies (eg, olpasiran, lepodisiran), and small oral molecules (eg, muvalaplin).⁷⁹

ASOs are synthetic short single-stranded DNA or RNA that regulate various genes and their variants by binding to complementary sequences. As pertains to Lp(a), ASOs bind to nuclear mRNA, activate ribonuclease-H, and lead to mRNA degradation, consequently inhibiting the production of apo(a) and ultimately lowering Lp(a) levels. APO(a)_Rx was the first ASO produced that targets Lp(a). In a phase 2 trial, it was well tolerated and led to a 78% reduction in Lp(a) levels.³³ This potency was further improved by covalently binding N-acetylgalactosamine (GalNAc) to enhance hepatocyte uptake of the drug, which led to pelacarsen, a second-generation ASO. This improvement enables pelacarsen to achieve the same Lp(a) reduction with a much lower dose.³³ Pelacarsen also lowers OxPL. In a phase 2 placebo-controlled trial in patients with established ASCVD and Lp(a) > 150 nmol/L, pelacarsen showed clinically meaningful and sustained reductions in Lp(a) of up to 80% compared with 6% seen in the placebo group.³³

siRNA-based therapies that interfere with Lp(a) synthesis include olpasiran and lepodisiran,^33,78 which are administered subcutaneously, well tolerated, and highly efficacious, reducing Lp(a) levels by up to 99%.^33,78 Olpasiran is a GalNAc-conjugated siRNA that degrades apo(a) mRNA using the RNA-induced silencing complex.⁷⁸ In the phase 2 OCEAN(a)-DOSE trial (Olpasiran trials of Cardiovascular Events And LipoproteiN[a] reduction–DOSE finding study), which included 281 patients with ASCVD and Lp(a) > 150 nmol/L, olpasiran demonstrated remarkable dose-dependent placebo-adjusted Lp(a) reductions at week 36 of 70.5% (10 mg every 12 weeks), 97.4% (75 mg every 12 weeks), 101.1% (225 mg every 12 weeks), and 100.5% (225 mg every 24 weeks), all P < .001.³³ The drug also modestly reduced LDL-C and apoB-100 concentrations, with adverse events similar to placebo except for more frequent hypersensitivity and injection-site reactions.³³ Lepodisiran is a GalNAc-conjugated, mixed 2’-O-methyl, 2’-fluoro and unmodified dicer siRNA that demonstrated dose-dependent and sustained Lp(a) lowering in healthy individuals with Lp(a) ≥ 75 nmol/L, with a well-tolerated safety profile.

Muvalaplin is an orally administered medication that inhibits Lp(a) formation through a novel mechanism by disrupting the noncovalent interaction between apo(a) and apoB-100, thereby preventing the disulfide bond formation required for Lp(a) assembly; it has been shown to lower Lp(a) by 50% to 86%.^78,79,80