Aortic valve calcification (AVC) is the most common valvular disease, affecting approximately 30% of the population over 65 years of age. The prevalence of calcific aortic valve stenosis (CAVS) exceeds 10% in people over 75 years of age, while severe disease affects more than 3% of the elderly [1]. Given the unfavorable prognosis of the disease and the lack of efficient non-invasive methods to modify the disease’s natural history, AVC and CAVS remain a major public health concern [2].

To date, no pharmaceutical agent has been shown to improve or slow progression of the disease, and the only interventions available for patients with severe and symptomatic CAVS are surgical aortic valve replacement (SAVR) and transcatheter aortic valve implantation (TAVI).

The recognition of lipoprotein(a) [Lp(a)] as a strong and potentially modifiable risk factor for the development of CAVS brought CAVS treatment to the forefront [3]. Lp(a) is associated with increased cardiovascular risk, including coronary artery disease (CAD), peripheral artery disease (PAD), heart failure, CAVS, and non-cardioembolic ischemic stroke ^[3][4][5][6][3,4,5,6]. Until recently, hypolipidemic agents had no or mild effect on Lp(a) levels. However, new emerging agents, targeting Lp(a) synthesis specifically, are showing significant effectiveness and have rekindled hope for a promising future in the treatment of patients with high Lp(a) levels. The effect of these agents on AVC and disease progression is yet to be investigated.

Particular research is being conducted on the correlation of CAVS or AVC with Lp(a). AVC, also known as aortic sclerosis, is defined as the non-obstructive calcification of the aortic valve (aortic jet velocity ≤ 2.5 m/s), and is considered to be a precursor of CAVS (aortic jet velocity > 2.5 m/s). The two entities are believed to be different stages of the same pathology. Both AVC and CAVS are characterized by an age-related incidence and an increasing prevalence in the general population because of population aging.

AVC and CAVS were believed to be the result of a passive, age-related process secondary to valvular damage, known as the ‘wear and tear’ mechanical stress theory. However, recent studies suggest that AVC is an active process, similar to atherosclerosis ^[7][15].

Lipid deposition, macrophage infiltration, extracellular matrix (ECM) dysregulation including collagen overproduction, and osteoblastic transitioning of valve interstitial cells (VICs) comprise the disease’s pathophysiology ^[7][15].

Age, male gender, low-density lipoprotein-cholesterol (LDL-C), arterial hypertension, and smoking are well described risk factors for the development of CAVS ^[8][16], while bicuspid aortic valve (BAV) is strongly associated with earlier disease development and faster disease progression ^[9][17]. Lp(a) constitutes a novel recognized risk factor for the development of AVC, with the first studies suggesting a potential association published in the late 1990s ^[8][10][16,18].

Currently, the crucial and causal role of Lp(a) in the AVC is well established, mainly via Mendelian randomization analyses conducted in general population studies. Combined data from two large prospective population studies, the Copenhagen City Heart Study and the Copenhagen General Population Study indicate a linear association of Lp(a) levels with the risk of a future diagnosis of valve calcification or stenosis. In particular, Lp(a) seems to be an independent risk factor for the development of AVC, even in patients with levels between the 67th and the 89th percentile (20 to 64 mg/dL) (1.6 [95% CI: 1.1 to 2.4]), while in those with levels above the 95th percentile (>90 mg/dL), an almost three-fold increased risk is observed ^[11][19].

Moreover, in the EPIC-NORFOLK prospective study, patients with Lp(a) > 50 mg/dL had a two-fold higher risk for the development AVC (2.15 (1.39–3.32), p = 0.001) ^[12][20].

The increased prevalence of AVC and CAVS in patients with familial hypercholesterolemia (FH) has also been investigated since the late 1990s ^[13][14][21,22]. At this time, the defining role of Lp(a) in disease prevalence and progression was yet to be established.

FH is a common autosomal (prevalence of heterozygous FH [heFH] 1 in 250], dominantly inherited disorder of cholesterol metabolism caused by mutations in the LDL receptor (LDLR) gene, the apolipoprotein-B (apoB) gene or the proprotein convertase subtilisin/kexin type 9 (PCSK9) gene. FH is characterized by elevated LDL-C levels which are almost two times higher in heterozygous subjects compared to the general population. In parallel, patients with FH seem to have elevated Lp(a) levels compared to non-FH subjects, potentially due to decreased Lp(a) clearance ^[15][23]. AVC seems to be almost two times more frequent in heFH patients compared to the general population, while non-LDLR mutations seem to be even more predictive of AVC compared with LDLR mutations. As both FH and high Lp(a) levels are strongly associated with an increased prevalence of AVC and aortic stenosis (AS), a two-shot hypothesis is suggested in those subjects ^[16][24].

Vongpromek et al. reported a significant association between Lp(a) levels and AVC in 129 asymptomatic heFH patients treated with statins. AVC was detected and quantified by computed tomography (CT) and Lp(a) levels > 50 mg/dL were an independent predictor of AVC. Interestingly, Lp(a) was not associated with coronary artery calcification (CAC), suggesting a different pathophysiology of these two entities ^[17][18][25,26]. This is also supported by a recent study showing that Lp(a) ≥ 70 mg/dL in 191 patients with stable CAD was specifically associated with the progression of low-attenuation and fibro-fatty plaques, but not with plaque calcification ^[19][27].

Bicuspid aortic valve (BAV), found in ~1% of the general population, is strongly associated with early AVC and AS. The majority of patients under 65 years old presenting with AS have a BAV, illustrating, therefore, the significance of aortic valve malformations in disease development ^[9][17].

The accelerated pattern of AVC in BAV offers a good model to investigate the association of AVC with Lp(a) levels. However, there is very little data on this field. A case-control study conducted by Sticchi et al., including 69 subjects with ultrasound evaluated BAV, demonstrated a statistically significant correlation between Lp(a) levels and AVC prevalence and severity in BAV patients ^[20][28].

Even if Lp(a) association with AVC is evident, its causative link to the disease remained in question until the determination of a specific single-nucleotide polymorphism (SNP) on the Lp(a) gene, namely the rs10455872 variant, which was associated with elevated Lp(a) levels as well as with the presence of clinically important AVC. In a study by Thanassoulis et al., the presence of the allele rs10455872 was associated with a two-fold increase in AVC detected by Computed Tomography (CT), even after adjustments for other known risk factors ^[21][29], suggesting a causal association of Lp(a) with the disease.

Elevated Lp(a) levels may be associated not only with disease incidence but also with faster disease progression and worse prognosis.

The ASTRONOMER trial constitutes a landmark on the research of Lp(a)-mediated AVC. Two hundred and twenty patients with mild to moderate CAVS were followed up using annual Doppler echocardiography and measurement of peak aortic valve velocity for a median period of 3.5 years. Patients in the top tertile of Lp(a) levels (>58.5 mg/dL) demonstrated faster disease progression compared to patients in the bottom two tertiles (<58.5 mg/dL) [peak aortic valve velocity: +0.26 ± 0.03 vs. +0.17 ± 0.02 m/s/year; p = 0.005], while they presented with a two-fold higher rate of clinical events, including Aortic Valve Replacement (AVR) and death. These endpoints were not affected by the presence of BAV or other AVC risk factors. In the same study, age is underlined as a determinant factor. Patients <57-year-old belonging to the top tertile had a twofold increase in disease progression compared with patients in the bottom tertile, while in those over 57 years old, Lp(a) contribution on CAVS progression was not statistically significant ^[22][14]. Even more, a secondary analysis of the ASTRONOMER trial suggests a linear association between Lp(a) levels and CAVS progression ^[23][30].

Another study, conducted by Zheng et al., using positron emission tomography (PET)-CT imaging and echocardiography follow-up and including data from two prospective multimodality imaging studies (SALTIRE and RING OF FIRE studies), came to confirm the ASTRONOMER trials’ conclusion on Lp(a)’s linear association with CAVS progression.

A recent study by Kaiser et al. came to contest the correlation of Lp(a) levels with AVC progression. 922 subjects recruited from the population-based Rotterdam Study were followed-up for a median period of 14 years using CT imaging. The results confirmed the association of baseline Lp(a) levels with the subsequent development of AVC in patients without AVC at baseline. However, in those with a preexisting AVC, disease progression was independent of Lp(a) levels. These results suggest that any emerging Lp(a) lowering treatment may be efficient only in the pre-sclerosis phase ^[24][32].

The mechanisms by which Lp(a) contributes to AVC and AS remain under investigation. AVC is believed to be a procedure similar to atherosclerosis. Inflammation, lipid deposition, fibrosis and calcification compose the histologic features observed in both entities. Calcification seems to be a key factor in both AVC and atherosclerosis. However, the pathophysiology of calcification differs between the two entities and AVC seems to be independent of atheroma calcification ^[25][26][33,34].

Calcification of the aortic valve is the result of a cell differentiation process. Specifically, VICs, the predominant population of cells in the aortic valves, are characterized by excessive plasticity and the potential to differentiate into osteoblast-like cells as a result of various stimuli. In vitro studies suggest a determinant role of Lp(a) on VIC transition toward osteoblast-like cells. Zheng et al. investigated the potential of Lp(a) in inducing osteogenic differentiation of VICs and concluded that one week of exposure of VICs in vitro to high Lp(a) levels leads to a significant increase in pro-inflammatory (interleukin-6 [IL-6]) and pro-osteoblastic transcription factors [bone morphogenetic protein-2 (BMP2), and runt-related transcription factor 2 (RUNX2)]. In addition, the effect of Lp(a) seems to be attenuated after its oxidized phospholipid (OxPL) content inactivation via either monoclonal antibodies targeting OxPL or a specific mutation on apolipoprotein(a) [Apo(a)] abrogating its ability to bind OxPL, thus highlighting a functional role of OxPL ^[27][31].

(a)

Lp(a) still preserves LDL properties and, thus, significantly contributes to intravalvular lipid deposition. Some studies propose that Lp(a) also has a role in the wound-healing process. Given the mechanical stress to which aortic valve leaflets are continuously submitted and the subsequent microfractures in their structure, Lp(a) may have an essential role in the lipid deposition process conducted via the healing process. Lipid deposition seems to contribute to VIC differentiation via the BMP2 pathway activation ^[26][34].

(b)

The OxPL content of apo(a) and its pro-inflammatory properties seem to be key factors of AVC. Inflammation is believed to lead to the activation of the calcification process via the activation of an innate immune response. OxPLs exhibit damage-associated molecular patterns (DAMPs) which, through toll-like receptors (TLRs) expressed on the VIC surface and the Nuclear factor kappa B (NF-kB) pathway, lead to the expression of IL-6. In vitro studies have demonstrated that IL-6 has the potential to activate the BMP2 pathway, thus leading to VIC differentiation ^[26][34].

(c)

Autotaxin, an oxidizing enzyme, may also have a crucial role in VIC transition. By binding to the Lp(a) molecule, autotaxin is transferred to aortic valve leaflets and triggers the oxidative transformation of phospholipid lysophosphatidylcholine (LysoPC) to lysophosphatidic acid (LysoPA). LysoPA promotes the expression of IL-6 and the activation of the BMP2 pathway leading, once again, to VIC differentiation into osteoblast-like cells, and finally to AVC ^[26][28][34,35].

The similarities of AS with the atherogenic process and the association of cholesterol with AS raised the expectation that an intense LLT with a strong statin would be able to reduce the rate of AS progression. This hypothesis was tested in the ASTRONOMER trial, which explored the effect of rosuvastatin 40 mg on CAVS progression in 269 asymptomatic patients with mild to moderate CAVS in recruitment time. All patients had annual echocardiographic assessment of AS progression for a median follow-up period of 3.5 years ^[29][36]. Rosuvastatin failed to positively affect the annualized increase in the peak AS gradient and these results were in accordance with those of a previous double-blind, placebo-controlled trial, which also failed to demonstrate any significant effect of atorvastatin on CAVS progression ^[30][37].

Statin treatment has also been associated with faster progression of calcification in atheromatous plaques. In the PARADIGM study, a prospective cohort study investigating the effect of statins on plaque characteristics, a faster plaque calcification process was observed in the statin group ^[31][38]. Given the similarities between atheromatosis and AVC, a comparable effect of statins on aortic valve leaflets has been proposed. In contrast with coronary artery plaques, where calcification progression seems to have a positive effect as it stabilizes the atheroma, a similar effect on the aortic valve could have a negative impact, increasing leaflet stiffness and promoting stenosis.

There are no currently available LLTs targeting only Lp(a) levels. Therefore, researchers present the effect of various LLTs, primarily LDL-C lowering therapies, on Lp(a):

1. Lipoprotein apheresis is the most efficient method for reducing Lp(a) levels that is available to date. Studies conducted in patients with FH who underwent LDL apheresis have shown a ~60% reduction of Lp(a) concentration ^[32][33][39,40]. These promising outcomes and the lack of equally sufficient non-invasive approaches explain the fact that in Germany, since 2008, Lp(a) apheresis has been available for patients with elevated Lp(a) levels and established cardiovascular disease (CVD), even without elevated LDL levels ^[34][41]. Even if Lp(a) reduction via lipoprotein apheresis seems to be essential, its impact on CVD risk remains under investigation. More useful data is expected from the MultiSELECt trial, a multicenter multinational prospective two-arm matched-pair observational study which is expected to be completed by the end of 2022 ^[35][42].

2. Statins have a controversial effect on Lp(a) levels. Despite the initial conviction that statin treatment could decrease Lp(a) levels by augmenting LDLR expression, various studies have demonstrated a neutral or, even, negative effect of statins on Lp(a) levels ^{[36][37][38][39][40][41][42]}[43,44,45,46,47,48,49]. A meta-analysis of 3 RCTs showed that statins increased Lp(a) levels by ~10% ^[41][48]. Moreover, different types and doses of statins have been proposed to have separate effects on Lp(a) levels. In particular, it has been suggested that atorvastatin has a slightly worse effect on Lp(a) levels compared to other statins, and its effect is dose-dependent, while the effect of pitavastatin seems to be more favorable ^[42][49].

3. PCSK9 inhibitors (PCSK9i) also reduce Lp(a) levels to a certain extent. In a systematic review and meta-analysis of phase II and phase III RCTs including evolocumab and alirocumab in comparison to placebo and/or ezetimibe and/or other LLT, a ~27% reduction of Lp(a) levels was observed ^[43][50]. At first, it was believed that Lp(a) reduction due to PCSK9i is achieved exclusively via LDLR up-regulation. However, a pooled analysis from 10 ODYSSEY Phase 3 studies indicated a high and unexplained discordance between LDL-C and Lp(a) reduction in patients treated with a PCSK9i. If LDLR up-regulation could fully explain Lp(a) reduction due to PCSK9i, then LDL-C and Lp(a) reduction should follow a similar pattern. The absence of such a pattern suggests the presence of additional mechanisms ^[44][51]. It is therefore believed that Lp(a) reduction due to PCSK9i administration is achieved by a dual mechanism including both increased Lp(a) catabolism and decreased Lp(a) production ^[45][52]. LDLR up-regulation could explain the increased catabolism of Lp(a) particles ^[46][53], while in contrast, the inhibition of Lp(a) expression due to PCSK9i is less understood ^[47][54].

4. Niacin and cholesteryl ester transfer protein (CETP) inhibitors have also been shown to decrease Lp(a) by ~20% and 30%, respectively ^[48][49][55,56]. However, both medications failed to show a clinical benefit in patients with CAD in the era of statins ^[50][51][57,58].

In conclusion, from currently used hypolipidemic drugs, only PCSK9i modestly decreases Lp(a). Analysis of the FOURIER Trial suggests that Lp(a) reduction by evolocumab possibly contributes to the clinical benefit. In particular, it was shown that individuals with higher baseline Lp(a) levels had greater absolute Lp(a) reductions with evolocumab, and this was associated with greater coronary risk reduction ^[52][59]. These findings are encouraging and somehow in contrast to a Mendelian randomization analysis which proposed that a very large Lp(a) decrease, i.e., ~100 mg/dL, is required to have a similar CVD risk reduction with an LDL-C reduction of 39 mg/dL ^[53][60].

A fundamentally novel approach is the use of agents that target RNA in order to interfere with the synthesis of Lp(a) through intracellular pathways, leading to gene silencing. Currently, there are two types of oligonucleotide-mediated gene silencing agents being investigated: “antisense” oligonucleotide-based molecules (ASOs) and small interfering RNAs (siRNAs). Both approaches act by creating an “antisense” oligonucleotide-based strand, which binds to its cognate RNA in vivo and inhibits apo(a) synthesis. Their main difference is that ASOs are single stranded, while siRNAs are double stranded, containing a sense and an antisense strand. The pharmacologically active half is the antisense strand, while the sense strand acts as a ‘‘drug delivery device’’ ^[54][61].

The concept of this treatment was first documented with the use of mipomersen, an FDA-approved ASO inhibitor of apoB-100 synthesis. Mipomersen acts by directly inhibiting apoB-100 synthesis and, thus, LDL plasma concentration. Mipomersen has been used in adults and pediatric populations with homozygous and heFH, already treated with maximally tolerated statin treatment or plasmapheresis ^[55][56][62,63]. A pooled analysis of 382 patients with FH or severe polygenic hypercholesterolemia demonstrated a significant 26% reduction of Lp(a) in the mipomersen arm ^[55][62]. The mechanism proposed by the authors is that the inhibition of apoB-100 synthesis results in limited apoB-100 particle availability for the formation of Lp(a). This hypothesis is also supported by non-human studies where mipomersen resulted in reduced apoB-100 containing particles while hepatic apo(a) mRNA and apo(a) plasma concentration remained stable ^[57][58][64,65].

The first specific ASO directly targeting Lp(a) synthesis, ASO 144367, was investigated in transgenic mouse models. The results were revolutionary, showing for the first time that a pharmaceutical agent can drastically reduce Lp(a) levels and their associated OxPL levels up to 86% and 93%, respectively ^[59][66].

Following the promising results from ASO 144367, multiple ASOs were screened for their ability to reduce apo(a) expression in vitro. ISIS-APO(a)Rx, also known as IONIS-APO(a)Rx, was identified as optimal for further investigation in non-human and human trials ^[59][66].

In a phase I clinical trial of IONIS-APO(a)Rx in healthy adults, a single-dose (50 mg, 100 mg, 200 mg, or 300 mg or placebo) as well as a multi-dose (100 mg, 200 mg, or 300 mg or placebo at days 1, 3, 5, 8, 15, and 22) subcutaneous injection schedule was followed. Single-dose administration did not reduce Lp(a) levels significantly. In contrast, multi-dose administration resulted in a dose-dependent Lp(a) reduction (100 mg 39.6%, p = 0.005 vs. placebo; 200 mg 59%, p = 0.001 vs. placebo; 300 mg 77.8%, p = 0.001 vs. placebo) on day 36. Lp(a) reduction was sustained almost 3 months later ^[60][67]. This positive effect was also demonstrated in the subsequent phase II clinical trial. The population of the study was divided into two cohorts based on baseline Lp(a) levels (125–437 nmol/L in cohort A; ≥438 nmol/L in cohort B). Following a multi-dose subcutaneous injection schedule (100 mg, 200 mg, and then 300 mg, once a week for 4 weeks each vs placebo) a significant reduction of Lp(a) levels was observed in both cohorts [66.8% (SD 20.6) in cohort A and 71.6% (SD 13.0) in cohort B (both p < 0.0001 vs. pooled placebo)] almost 3 months later ^[61][62][68,69].

In order to optimize ASO hepatocyte endocytosis, IONIS APO(a)-LRx, an ASO conjugated with N-acetylgalactosamine (GalNAc), was designed. GalNAc is a molecule that binds to the asialoglycoprotein receptor and that is highly expressed on hepatocytes, resulting in rapid endocytosis and therefore improving drug potency and tolerability. IONIS APO(a)-LRx was, therefore, investigated in a phase 1/2a first-in-man trial using lower doses compared to IONIS APO-Rx trials, and resulting in even better effectiveness (66% in the 10 mg group, 80% in the 20 mg group, and 92% in the 40 mg group (p = 0.0007 for all vs. placebo) at day 36. Treatment with both IONIS-APO(a)Rx and IONIS-APO(a)-LRx, significantly reduced OxPL and OxPL-apoB, proinflammatory particles that promote atherogenesis as well as calcification of the aortic valve ^[61][62][68,69].

Lastly, an ongoing phase III randomized, placebo-controlled clinical trial (Lp(a)HORIZON Trial) will assess the effect of IONIS-APO(a)-LRx (pelacarsen) on major cardiovascular events in patients with established CVD and Lp(a) ≥70 mg/dL. Primary endpoints include time to first occurrence of major adverse cardiovascular events. The trial currently has 8221 patients participating, and it is expected to end in 2025 ^[63][70].

Following ASOs, a second emerging intervention directly targeting Lp(a) production is the use of siRNAs in order to inhibit apo(a) mRNA expression. Studies in non-human primates showed a robust reduction of plasma Lp(a), over 95%, after a single dose of the agent, which was sustained over a 3-week follow-up period ^[64][71]. Subsequent human phase I studies were conducted in patients with elevated Lp(a) levels and without known CVD. SLN360, an siRNA targeting apo(a) mRNA linked with GalNac, was used in ascending doses, and a dose-dependent and persisting (over 150 days) reduction of Lp(a), up to 98% was demonstrated ^[65][72]. SLN360 is currently investigated in a multi-center, randomized, double-blind, placebo-controlled, phase II study, expected to be completed by the end of 2024 ^[66][73].

6. Novel Lp(a)-Lowering Therapies: The Dawn of Pharmaceutical Treatment of AS?

The well documented association of Lp(a) with the fibro-calcific remodeling of the aortic valve suggests that Lp(a) might be a good target to prevent AVS. Therefore, it is tempting to speculate that the novel Lp(a)-lowering therapies might be able to prevent AVC or to decelerate the progression of AVC to CAVS. However, there is currently no randomized controlled trial testing this hypothesis, and there are several issues that arise under this perspective:

(1)

What is the best technique to assess the effectiveness of Lp(a)-lowering agents? Echocardiography is an excellent tool to assess AVS, but it is unable to quantitate AVC. On the contrary, CT can measure AVC and can objectively evaluate the effectiveness of the Lp(a)-lowering treatments which will possibly exert their beneficial effect by delaying the calcification process.

(2)

Lp(a)-lowering agents should be ideally tested after the completion of phase III trials which explore their safety and effectiveness on cardiovascular events.

(3)

How early should Lp(a)-lowering agents be given? “The earlier, the better” principle of statins administration in acute coronary syndrome is likely to be applicable in the setting of AVC. There is data suggesting that the impact of Lp(a) on the progression of calcification may be weak once AVC has been initiated ^[19][27].

(4)

What is the target population that is more likely to benefit from the new pharmaceutical interventions? It is appropriate to test the new Lp(a)-lowering therapies in populations at high risk of developing AVC/CAVS, i.e., patients with high Lp(a) levels and BAV, or possibly heFH patients? Particularly, in the case of heFH, by lowering Lp(a) it is possible to obtain a dual beneficial effect, i.e., a reduction of cardiovascular events and the prevention of AVC/CAVS.

(5)

What will be the threshold Lp(a) levels which will justify a pharmaceutical intervention? In the Lp(a)HORIZON trial, pelacarsen is tested in patients with established CVD and Lp(a) ≥ 70 mg/dL. For the prevention of AVC/CAVS it might be sensible to set higher Lp(a) cutoff levels.

All of these issues comprise a complex puzzle towards the future perspective for the conservative treatment of AVC/CAVS. The topic is getting more complicated by considering the complex pathophysiology of AVC/CAVS, where Lp(a) is only one significant player.