Calcific aortic valve disease (CAVD) is one of the most common valve disorders (1). CAVD is characterized by calcification and remodeling of the valve leaflets, which often progresses to aortic sclerosis and stenosis, eventually leading to heart failure, angina, death, and other serious adverse cardiovascular events (2, 3). 2020 VHD guideline recommended the intervention of symptomatic AVS mainly apply to SAVR and TAVI. Surgical treatment is not performed routinely in asymptomatic patients (4). With more and more in-depth studies of pathogenesis, researchers are exploring targeted drugs to delay disease progression (5). A meta-analysis included three RCT and five observational studies to analyze the efficacy of ACEI/ARB to CAVD (6), the results showed that there was no statistically significant difference in all-cause mortality between the two groups, but the AVR rate of the treatment group was lower than that of the control group, which needed large-scale RCT to prove. Multiple rigorous RCTs have shown negative efficacy of statins (7–9). Besides, researchers also explored the targets on phosphate/calcium-metabolism and nitric oxide and IGF-1 signaling pathway, which need to be further proved (5).

Globally, about 10–30% of the population has high lp(a) levels ≥50 mg/dl (10). Epidemiological and genetic evidence suggested that high lp(a) concentration had a direct relationship with cardiovascular disease (11). Previous cytological studies have shown that lipoprotein(a) [lp(a)] played an important role in the pathogenesis of CAVD through increasing inflammation and oxidative stress, promoting calcium deposition of valvular interstitial cells (VICs) (12, 13). Mendelian randomization studies suggested that LPA genotype, which could mediate the levels of lp(a), had a strong relationship with CAVD (14). Multiple genome-wide association studies had shown that the rs10455872 genetic variant in LPA, which was associated with higher lp(a) levels, was independently related to an augmented risk of CAVD (15, 16). However, evidence from clinical studies was inconsistent. Some studies indicated comparing to those without CAVD, patients with CAVD had significantly higher levels of lp(a) (17, 18), while some studies suggested there were no statistical differences in lp(a) levels between CAVD and controls (19). In response to the lack of systematic and comprehensive evidence on the clinical association between lp(a) and CAVD. We performed an extensive systematic review and meta-analysis to assess whether elevated lp(a) significantly affected the incidence of CAVD.

The systematic review and meta-analysis were performed following the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) (20) and Meta-analyses of Observational Studies in Epidemiology (MOOSE) checklist (21). The protocol has been registered in PROSPERO database (registration number: CRD42021273149).

We evaluated the relationship between elevated plasma lp(a) level and CAVD. The major findings were as follows: (1) The incidence of CAVD was higher in the high lp(a) level group than in the low-level group, as reported by most of the original studies. (2) The pooled results of lp(a) 50 mg/dL group supported that plasma lp(a) ≥50 mg/dL might be a risk factor for CAVD. Whereas, there was insufficient evidence for the association between plasma lp(a) ≥30 mg/dL and CAVD. Also, substantial heterogeneity between studies was not well-explained by subgroup analysis with a 30 mg/dL lp(a) cut-off value, which called for caution when interpreting the result. (3) Patients with plasma lp(a) ≥50 mg/dL might have a higher risk of CAVD than those with lp(a) ≥30 mg/dL. Therefore, we speculated that the magnitude of lp(a) concentration might have a dose-response relationship with CAVD. (4) The progression of CAVD has multiple stages, including calcification and stenosis, etc. AVC, an independent predictor of cardiovascular events (42), not only may progress to AVS (43), but increases the risk of CAD and all-cause mortality (44, 45). Severe AVS directly correlates with heart failure and death (2). Therefore, the lp(a) effect on AVC and AVS were analyzed separately. The results showed that high plasma lp(a) (≥30 mg/dL and ≥50 mg/dL) was associated with AVC, the relationship of AVS with lp(a) ≥50 mg/dL was significant but not with lp(a) ≥30 mg/dL. (5) The subgroup analysis by study types in lp(a) 30 mg/dL group concluded a positive finding in the cohort studies instead of case-control studies. Included case-control studies have the disadvantages of small sample size and the inherent nature of recall and select bias, which might cause inaccurate reporting of the results (46). And due to the larger sample size, the cohort studies had narrower confidence intervals and higher weight with more accurate estimates. Therefore, large sample cohort studies are needed to further explore the relationship between lp(a) ≥30 mg/dL and CAVD. The analysis result in Europe was more significant than in America, and no meaningful association of lp(a) ≥30 mg/dL was found in America. The participants of two American studies mainly involved Asians, Hispanics, Caucasians and Blacks, and the initial results were inconsistent among different races. However, the race of European studies was relatively homogeneous. Therefore, there are possible ethnic effect factors in the pathogenesis of lp(a)-mediated CAVD (27, 40). And a research article from the MESA study pointed out a possible race/ethnicity-related modification of lp(a) and coronary heart disease events (24). Their findings suggested that the 30 mg/dL cutoff for lp(a) is inappropriate in Caucasian and Hispanic individuals, and the higher 50 mg/dL cutoff should be considered. In contrast, the 30 mg/dL cutoff remains suitable in Black individuals. From the present evidence, the cut-off values of lp(a) differed according to different race/ethnicity groups. But most studies agreed on the cut-off values of 30 and 50 mg/dL. That's why we analyzed using both cut-off values in our meta-analysis. Our subgroup analysis results agree with the findings from Guan et al. (24). (6) The Egger's test and the Begg's test had certified that there was no publication bias in these studies. The number of included studies was <10 and the estimates we used were inherently related to their standard errors, leading to relatively imprecise test power of the funnel plot. Whereas, it was undeniable that selection bias and true heterogeneity might also be related to the asymmetry, such as unavailable unpublished studies, which inevitably lead to publication bias. And we conducted the subgroup analysis to investigate the source of heterogeneity, and this deficiency was made up to some extent.

Besides other conventional risk factors like age, sex, hypertension, and smoking, dyslipidemia played a vital role in calcific aortic valve disease (30). Lipids oxidation may promote chronic low-grade inflammation in the aortic valve, which frequently induces the osteogenic process of VICs (47–49). However, whether LDL-C or HDL-C has casual associations with CAVD remains controversial (50, 51). And two meta-analyses about the effect of statins on aortic valve stenosis demonstrated statins might not retard the progression of valve stenosis even though LDL-C concentration was reduced (52, 53). Instead, attention has turned to lp(a). A rapidly-growing body of evidence demonstrated that lp(a) has a bright prospect in predicting and treating CAVD. A previous systematic review speculated plasma lp(a) might be related to the occurrence and progression of AVS. However, this study did not perform a quantitative meta-analysis (54). Our work, which has been thoroughly searched, rigorously screened and quantitatively analyzed, found that lp(a) ≥50 mg/dL could potentially be a proper risk factor for CAVD. Therefore, lp(a) concentration could be more suitable for assessing the incidence of CAVD, compared with that of LDL-C and HDL-C.

Several plausible mechanisms may account for the underlying pathophysiology of lp(a)-mediated CAVD. Lp(a) is a low-density lipoprotein-like structure, which mainly contains apolipoprotein B (apo B) covalently bound to apolipoprotein(a) [apo(a)] (55, 56), transporting some pro-inflammatory and pro-osteogenic mediators. When aortic valve leaflets are damaged to mechanical or shear stress, excessive lp(a) could infiltrate and accumulate the valve to stimulate inflammation, calcification, and fibrosis (57–59). Oxidized phospholipids (OxPL), primarily carried by lp(a) complexes, has been demonstrated the association with CAVD (60, 61). Zheng et al. have revealed that the high content of lp(a) and OxPL-apoB was independently associated with increased active tissue calcification and clinical events such as AVR or all-cause mortality (12). OxPL might be hydrolyzed to lysophosphatidylcholine (LPC) in the presence of phospholipases, promoting valvular inflammation, thickening and mineralization (62). Lipoprotein-associated phospholipase A2 (Lp-PLA2) and autotaxin (ATX) might be the key phospholipases in the metabolism of lp(a)-OxPL. Lp-PLA2, with a high affinity to lp(a), could convert OxPL into LPC, increasing the expression of phosphate-related genes (63, 64). ATX, mainly combined with apo(a) of lp(a) particles, could transform LPC into LysoPA, which drives the inflammatory and osteogenic program through the NF-κB/IL-6/BMP pathway (41, 65). In addition, lp(a) could directly mediate remodeling and calcification of VICs through inducing MAPKs signaling pathway and the expression of pro-osteogenic factors like bone-specific transcription factor SP7 (osterix), BMP-2, and BMP-4 (66).

Welsh et al. revealed reducing the baseline lp(a) levels by 80%, patients with lp(a)≥175 nmol/l and baseline CVD could decrease the risk of CVD by 20% (67). And the conclusions of this study suggested that lowering lp(a) levels might be an important way to intervene CAVD. Pharmacotherapies of lp(a) include PCSK9 inhibitor and nucleic acid antisense, etc. In the FOURIER study, it was found that PCSK9 inhibitors could reduce the plasma Lp(a) level on average 26.9%, and patients with higher Lp(a) levels had a 23% lower risk of major cardiovascular events (68). In the ODYSSEY OUTCOMES trial, the reduction of Lp(a) levels by PCSK9 inhibitors was associated with a decrease in the risk of cardiovascular disease (69). Mipomersen is a 2′-O-methoxyethyl modified second-generation antisense oligonucleotide, which binds to homologous Apo B messenger RNA to inhibit the synthesis of ApoB-100. It can significantly reduce Apo B and Lp(a) levels (70). IONIS-APO(a)Rx, an oligonucleotide targeting lp(a), could reduce lp(a) levels by 66–92%, OxPL-ApoB and OxPL-apo(a) decreased moderately, which play a vital role in pathogenesis of CAVD (71).

There are several strengths of our meta-analysis. Firstly, this is the first meta-analysis assessing the association between lp(a) and CAVD. Secondly, our analysis included a comprehensive search strategy, a considerable number of participants, a subgroup analysis for heterogeneity. And we integrated all relevant cross-sectional and case-control studies, allowing us to determine whether there was a significant relationship between lp(a) and CAVD. Thirdly, most of the included studies have controlled the confounders, which ensured the reliability of the outcomes. Finally, we investigated the relationship between high lp(a) levels (≥50 and ≥30 mg/dL) and the incidence of CAVD, respectively, which provided some guidance for patients with elevated plasma lp(a) levels to assess their risk of CAVD. However, heterogeneity was observed in our meta-analysis, even though most of studies had adjusted multiple potential confounders. Therefore, the results of the present analyses should be interpreted with caution.

Based on the current cohort studies and case-control studies, we could conclude that patients with lp(a) ≥50 mg/dL were at a significantly high risk of CAVD. However, further large-scale prospective cohort studies with high quality and adequate control for confounders are necessary to draw a firm conclusion on the causality between plasma lp(a) and CAVD.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

QL and JJ conceived the study. QL and YY searched the studies, screened the studies, analyzed the data, and wrote the manuscript. QL and RL extracted the data. TW, YF, and RX assessed the risk of bias. JL, JJ, and HX reviewed and revised the manuscript. All authors contributed to the article and approved the submitted version.

This research was supported by grant CI2021A00917 from the China Academy of Chinese Medical Sciences Innovation Fund, grant ZYYCXTD-C-202007 from the Innovation Team and Talents Cultivation Program of National Administration of Traditional Chinese Medicine, and grant ZZ13-YQ-017 from the Central Public Welfare Research Institutes of China Academy of Chinese Medical Sciences.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.