Background
Bicuspid aortic valve (BAV) is the most common congenital defect characterized by two leaflets in the aortic position, which may have a spectrum of clinical presentations [
1,
2]. In terms of aortic stenosis, some interesting distinctions have been observed in BAV patients. Compared with normally formed tricuspid aortic valve (TAV), BAV patients often develop more severe stenosis with heavier calcification and/or fibrosis in aortic leaflets at a younger age [
3]. The difference is not limited to the valve itself, but also seems to involve the left ventricle (LV). Patients with BAV were reported to undergo less pronounced reverse LV remodeling than TAV [
4] after definitive treatment by means of surgical or transcatheter aortic valve replacement (TAVR). Despite these observations indicating bicuspid aortic stenosis to have its own characteristics, current follow-up and treatment of bicuspid aortic stenosis do not differ from TAV patients.
A handful of studies have indeed identified some intrinsic genetic bases engaged in the etiology of BAV, such as the gene of ACTA2 [
5] and NOTCH1 [
6]. However, as the final stage in processing genomic information, metabolites were considered to significantly influence cellular activity and directly reflect extracellular microenvironment [
7,
8]. Thus, metabolomics profiling technologies [
9] are more likely to determine novel disease biomarkers and provide insights in biological mechanisms, which would be essential to improve treatment strategies. To date, few studies have focused on the specific metabolic mechanisms involved in bicuspid aortic stenosis, especially after aortic valve replacement.
In this prospective study, we sought to incorporate metabolomics profiling technologies in the setting of TAVR, with patients of both bicuspid and tricuspid morphology. In particular, detailed analyses are planned for BAV patients to explore potential metabolomic differences in subgroups with worse baseline hemodynamics and poorer LV recovery post-TAVR.
Methods
Study population
From November 2015 to August 2016, patients with severe symptomatic aortic stenosis, who successfully underwent TAVR in our center with both pre- and post-procedural blood samples collected, were enrolled in the study, while patients who received TAVR due to failed bioprostheses were excluded (
N = 57, 32 females). Patients were divided into BAV and TAV groups according to their valve morphology confirmed and classified on multi-slice computed tomography (MSCT). TAVR was performed by the same heart team from the transfemoral access under general anesthesia, as described previously [
10]. All enrolled patients received the same management process and were treated with the same intervening protocol. Participants enrolled in this study signed the informed consent in accordance with the Declaration of Helsinki. This study was approved by the Institutional Review Board of West China Hospital, Sichuan University.
Echocardiography follow-up
Transthoracic echocardiography (TTE) were arranged before discharge and at routine follow-ups at the 1st month, 3rd month, 6
thmonth and 1st year after TAVR. To assess reverse LV remodeling after TAVR, LV mass (LVM) and LVM index (LVMI) were calculated [
11]. The change in LVM and LVMI was calculated as the value at follow-ups subtracted the value at baseline (shown as ∆LVM and ∆LVMI) and served as a surrogate endpoint to LV recovery post-TAVR.
Blood sample collection and processing
Blood sample was obtained from peripheral vein before valve deployment during the index TAVR procedure and at the 7th day after TAVR (normally the day of discharge) in K2EDTA-treated tubes (BD). Subsequently, these blood samples were centrifuged at 2000 g for 10 min to pellet the cellular elements. The supernatant plasma was stored at − 80 °C until sample preparation for liquid chromatography-mass spectrometry (LC-MS) and gas chromatography-mass spectrometry (GC-MS) analysis.
Metabolomics profiling was detected and processed by Chinese Academy of Sciences in Dalian based on methods described in previous studies [
12,
13]. Briefly, metabolomic analysis was conducted on ACQUITY Ultra Performance Liquid Chromatography (UPLC, Waters Corporation, Manchester, UK) system with ACQUITY UPLC BEH C8 1.7 μm (2.1 × 100 mm) column (Waters, Milford, MA) for positive ion metabolites separation setting at 50 °C and ACQUITY UPLC HSS T3 1.8 μm (2.1 × 100 mm) column (Waters, Milford, MA) for negative ion metabolites separation setting at 50 °C. A QP 2010 Plus GC-MS system (Shimadzu, Japan) with a DB-5MS (30 m × 250 μm × 0.25 μm, Agilent Technologies, USA) was used in GC-MS analysis. The procedures were performed in accordance with the manufacturers’ protocols of all devices. Then, raw data were converted to mzData formats via Agilent MassHunter Qualitative software (Agilent, Santa Clara, CA, USA). The program XCMS (version 1.40.0) (
https://xcmsonline.scripps.edu/) was used to preprocess the raw data, with the default parameters. In the quality control (QC) process, nearly 100% features in QC samples had relative standard deviation (RSD) distribution less than 30%, and were within 3SD, which demonstrated a satisfactory data quality. The resulting matrix was constructed by retention time, mass-to-charge ratio (m/z), and normalized ion intensities.
The model of partial least squares discriminant analysis (PLS-DA) was used to classify metabolic characteristics of bicuspid aortic stenosis from tricuspid aortic stenosis. The variable importance project (VIP) from PLS-DA was also calculated. Furthermore,
p-value from t-test analysis and fold-change value (FC) were executed to discover differentially expressed metabolites, which were exhibited by the volcano plot. With a pre-specified cutoff of
p-value< 0.05, FC > 1.2 and VIP > 1.2, the dominant molecules were to be identified. These analyses were completed with an in-house script by R platform (
https://www.r-project.org).
To futher identify the essential metabolic pathways and hub metabolites engaged in phenotypic alterations, pathway enrichment analysis was employed to summarize more interpretable results [
14]. However, if the number of metabolites identified above was over 100, weighted gene co-expression network analysis (WGCNA) algorithm was to be first utilized to narrow down the number of metabolites sent to pathway enrichment analysis [
15]. This protocol enabled the detection of hub molecules which were also correlated with clinical traits. Metabolites to be sent to patheway enrichment analysis were then selected by Venn diagram of both hub molecules defined by the WGCNA protocol and differentially expressed metabolites defined by the volcano plot. Pathway enrichment analysis was conducted with MetaboAnalyst platform (http://www.metaboanalyst.ca) [
16].
Subgroup analysis
For the subgroup analysis to explore the specific metabolites related to worse hemodynamics in BAV patients before TAVR, 4 m/s < maximal transaortic velocity (Vmax) < 4.5 m/s was defined as the lower Vmax group, whereas Vmax > 5 m/s was defined as the higher Vmax group. Post-procedurally, the value of ∆LVMI at 1 year was divided into quartiles. BAV patients in the 4th quartile were regarded to have the worst LV recovery, while those in the 1st quartile were regarded to have the best LV recovery.
Statistical analysis
Normality was assessed for all clinical datasets by the Shapiro-Wilk’s test. Continuous results were presented as the mean ± SD or median (IQR) according to their distribution. Independent t-test (for variables that were normally distributed) or Mann-Whitney U-test (for variables that were not normally distributed) was carried out to assess differences in clinical features between individuals in two groups. A two-tailed p value < 0.05 was considered as significant.
Discussion
The incorporation of metabolomics profiling technologies to TAVR patients offered us a unique way to explore bicuspid aortic stenosis in comparison with the normal tricuspid morphology, both before and after the stenotic status been corrected. The two major findings of this study are (1) BAV patients had an altered arginine and proline metabolism pathway compared with TAV, which sustained post-TAVR; (2) the arachidonic acid metabolism pathway seems to be associated with more severely affected hemodynamics, as well as poorly-recovered LV post-TAVR in BAV. These preliminary results provide novel insights to further elucidate the different disease spectrum of bicuspid aortic stenosis and determine possible biomarkers for monitoring progression and prognosis of bicuspid aortic stenosis.
Patients with bicuspid aortic stenosis are known to require aortic valve replacement at a much younger age than their tricuspid counterparts, usually within productive years of life [
17]. As we currently do not have means to effectively postpone the disease progress of aortic stenosis, determining the difference between bicuspid and tricuspid aortic stenosis may be an alternative way for BAV patients to at least face this inevitable intervention later in their life. Elevated arginine and proline metabolism pathway was found in TAV compared with BAV, regardless whether severe aortic stenosis was corrected or not. This finding suggested that metabolites in the arginine and proline pathway may be potential targets to delay the progression of bicuspid aortic stenosis. The arginine and proline metabolism pathway generates nitric oxide with Nitric Oxide Synthase (NOS) [
18,
19]. Elevated endothelial nitric oxide synthase (eNOS) has been found in TAV [
20], but eNOS was low expressed and irregularly distributed in BAV [
21]. ENOS is essential in regulating biological homeostasis. Limited eNOS reduces cellular NO level which could lead to oxidative stress, and promote oxidative damage [
22,
23]. Besides key enzymes in arginine and proline metabolism pathway, the detected crucial amino acids such as glutamine and proline, likewise, have biological significance for bicuspid aortic stenosis. Proline was previously reported to be the balancer of over oxidation, which was engaged in stabilizing proteins and antioxidant enzymes, and directly scavenging ROS [
24]. A study has also demonstrated that glutamine effectively controlled systemical inflammatory reactions, especially the molecules of IL-1, IL-6, IL-8, and IL-10 [
25], which might be the proposed factors for accelerating valve stenosis [
26]. Thus, targeting eNOS or the key amino acids in arginine and proline metabolism pathway might be potential targets to delay the progression of bicuspid aortic stenosis. Previously, perturbations in fatty acid metabolism has been identified in calcific coronary artery disease stratified by the severity of calcification [
27,
28]. As a pathology with some similar characteristics, calcific aortic valve disease might share these alterations. However, in this analysis, BAV patients did not achieve a statistically higher calcium burden on the leaflets than TAV patients and the coronary calcium score was not assessed, making it difficult to evaluate the impact of total calcium burden on metabolomic profiling.
In the BAV subgroup analysis, both more severely affected baseline hemodynamics and poorly recovered LV post-TAVR were linked to elevated arachidonic acid metabolism pathway. As an important pathway for inflammation, the alleviative arachidonic acid metabolism might be crucial to halt the pathologic progression of aortic stenosis, especially for BAV patients [
29,
30]. Within this pathway, the key enzyme of COX-2 was considered as an important drive for aortic stenosis by increasing valvular calcification and enhancing osteogenic genes including OPN and Runx2 [
29]. In addition, COX-2 was reported to be central in the AngII-induced macrophage recruitment and the expression of TNF-α of endothelial cells, which blockades macrophage infiltration and exaggerates expression of pro-inflammatory cytokines [
30]. Although there was no research reporting the role of arachidonic acid metabolism pathway in the recovery phase after aortic valve replacement, it was otherwise demonstrated to influence myocardial functions. Targeting metabolites such as cyclooxygenases and leukotrienes B4 in the arachidonic acid metabolism pathway protected against heart failure, decreased myocardial fibrosis and rescued myocardial function [
31,
32]. The involved mechanisms might lie in preserving microcirculation and limiting inflammation, such as inhibiting TNF-α-mediated pathway and inflammatory mediators [
33,
34]. Thus, targeting arachidonic acid metabolism pathway might be of tremendous therapeutic value in clinical practice and merits further detailed researches.
There were several limitations in this study. Firstly, blood samples were sent for metabolomics detection without particular filtering. Secondly, there was gender imbalance possibly by coincidence at the time of this study, but cautions should be taken when interpreting the difference of metabolomic profiling between the two valve morphologies due to this selection bias. Thirdly, it was difficult to draw definitive conclusions due to relatively small sample size without a validation cohort and limited related researches in this area. A larger metabolomics study combining validation experiments should be arranged to verify our findings and explore the detailed mechanisms of molecules involved in bicuspid aortic stenosis.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.