Abstract
Background:
Hypertrophic cardiomyopathy (HCM) is a heterogeneous condition that can lead to atrial fibrillation (AF), heart failure (HF) and sudden cardiac death (SCD) in many individuals but mild clinical impact in others. The mechanisms underlying this phenotypic heterogeneity are not well defined. The aim of this study was to use plasma proteomic profiling to help illuminate biomarkers that reflect or inform the heterogeneity observed in HCM.
Methods:
The OLINK antibody-based proteomic platform was used to measure plasma proteins in patients with genotype positive (sarcomeric) HCM participating in the Hypertrophic Cardiomyopathy Registry (HCMR). We assessed associations between plasma protein levels with clinical features, cardiac magnetic resonance (CMR) imaging metrics and development of atrial fibrillation.
Results:
We measured 275 proteins in 701 patients with sarcomeric HCM. There were associations between late gadolinium enhancement (LGE) with proteins reflecting neurohormonal activation (N-terminal brain natriuretic peptide and Angiotensin Converting Enzyme 2). Metrics of left ventricular remodeling had novel associations with proteins involved in vascular development and homeostasis (Vascular Endothelial Growth Factor-D and Thrombomodulin). Assessing clinical features, the European Society of Cardiology sudden cardiac death risk score was inversely associated with Stem Cell Factor. Incident AF was associated with mediators of inflammation and fibrosis (Matrix metallopeptidase 2 and Spondin 1).
Conclusions:
Proteomic profiling of sarcomeric HCM identified biomarkers associated with adverse imaging and clinical phenotypes. These circulating proteins are part of both established pathways, including neurohormonal activation and fibrosis, and less familiar pathways, including endothelial function and inflammatory proteins less well characterized in HCM. These findings highlight the value of plasma profiling to identify biomarkers of risk and to gain further insights into the pathophysiology of HCM.
Keywords: hypertrophic cardiomyopathy, proteomics, biomarkers
Introduction
Hypertrophic cardiomyopathy (HCM) affects roughly 1 in 500 adults and a causal sarcomere gene variant can be identified 30–60% of the time (1). The clinical outlook of HCM patients is remarkably heterogeneous, ranging from asymptomatic individuals to those who experience advanced heart failure (HF) or sudden cardiac death (SCD). The mechanisms underlying this heterogeneity are not well understood, thus hampering our ability to robustly predict risk and provide personalized clinical management. Proteomics, the large-scale study of proteins and their functions, has emerged as a valuable tool to improve risk stratification and to help unravel the underpinnings of cardiovascular disease (2)(3). Prior efforts in proteomic profiling in HCM have illuminated biological pathways altered in this disease. However, these studies have primarily used cardiac tissue from patients undergoing septal myectomy or transplantation (4–6). While proteomic profiling of cardiac tissue in HCM has the advantage of ensuring proteins of interest are being expressed in the heart, there are significant limitations to this approach. Profiling of myocardial biopsies which are taken at the time of surgery in individuals with advanced disease, limits both the sample size and generalizability of the studies to the HCM population at large given the bias in findings towards a sicker patient population.
The advent of circulating blood-based proteomics enables large scale population profiling and opportunities for risk stratification. Notably, proteomic profiling using plasma from participants in large population-based cohorts has demonstrated novel associations in other cardiovascular conditions, such as heart failure (7) and coronary artery disease (8). These studies have provided insight into potential biomarkers of disease and underlying mechanisms. While recent plasma proteomic studies have been applied to HCM, they have been conducted on cohorts with limited genotypic and phenotypic characterization and so have been limited in their ability to address the heterogeneity within HCM (9). Non-sarcomeric HCM is often a polygenic trait and characterized by an increased burden of cardiovascular comorbidities, including hypertension, compared to sarcomeric HCM (10). As these cardiometabolic risk factors may potentially contribute to the pathophysiology of HCM, identifying signatures of heterogeneity specific to HCM may prove to be challenging due to these confounding factors. This underscores the importance of proteomic studies in sarcomere mutation carriers.
In this study, we utilized the Olink platform to perform proteomic profiling in patients with sarcomeric HCM from the NHLBI-funded Hypertrophic Cardiomyopathy Registry (HCMR) (11). The HCMR cohort has been systematically genotyped and phenotyped with clinical risk assessment, echocardiography, and comprehensive cardiac magnetic resonance imaging (CMR) analysis. By leveraging proteomic techniques in a large cohort of patients with sarcomeric HCM, we aimed to identify biomarkers of adverse cardiac features in HCM to gain a deeper understanding of the complex manifestations and heterogeneity characteristic of this disease.
Methods
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Study Population
HCMR enrolled 2,755 patients with hypertrophic cardiomyopathy (HCM) into a prospective registry study. The study collected demographic, medical history, and clinically-performed echocardiographic data. Study-protocol comprehensive contrast-enhanced cardiac magnetic resonance (CMR) studies were performed and adjudicated by dedicated core laboratories in addition to genetic and biomarker analyses. The study population has been previously described (10). Patients have been prospectively followed by medical review for the composite primary outcome of cardiac death including SCD, aborted SCD, heart transplantation and left ventricular assist device implantation. Secondary outcomes have also been prospectively assessed, including all-cause mortality, ventricular tachyarrhythmia, heart failure hospitalization, stroke, and clinically important atrial fibrillation including events leading to electrical cardioversion, catheter ablation, hospitalization >24 hours, or clinical decisions to accept permanent AF. Written informed consent was obtained from all patients, and the study protocol was approved by the institutional review boards of the participating sites.
Genetics
Genetic analysis was performed at the University of Oxford, where sarcomere variant status was determined using amplicon-based sequencing on 36 genes associated with cardiomyopathy using the Illumina MiSeq platform. Bioinformatics analysis was conducted according to best practice guidelines using the Genome Analysis Toolkit. Data from publicly available resources (ClinVar [version 20190211] and gnomAD r2.1) and the Oxford Regional Genetics Laboratory in-house mutation database were used to inform variant classification. Based on gene-specific interpretation of variants, participants were dichotomized as ‘sarcomere mutation positive’ (individuals who carried pathogenic, likely pathogenic, and depending on the gene, a subset of variants of unknown significance with features that favored pathogenicity) and ‘sarcomere mutation negative’(no clinically significant sarcomere variant identified; non-sarcomeric HCM (10); Supplementary Methods). This study included only sarcomere mutation positive participants, herein referred to as sarcomeric HCM.
OLINK Proteomic Profiling
The Olink® Target 96 Cardiometabolic, Cardiovascular II & III panels (Olink Proteomics AB, Uppsala, Sweden) were used to measure proteins according to the manufacturer’s instructions. The Olink protocol utilizes the Proximity Extension Assay (PEA) technology, which allows for the simultaneous analysis of 92 analytes using only 1 μL of each sample. This technology has been well described previously (12). Briefly, pairs of oligonucleotide-labeled antibody probes specifically bind to their respective target proteins. When these probes come close to each other, the oligonucleotides hybridize, and a proximity-dependent DNA polymerization occurs. This generates a unique PCR target sequence. The resulting DNA sequence is then detected and quantified using a microfluidic real-time PCR instrument (Biomark HD, Fluidigm). To ensure data quality and account for variation between runs, internal controls including an extension control and an inter-plate control are used for normalization. The final output of the assay is presented as Normalized Protein eXpression (NPX) values, which are on a log2-scale. Higher NPX values indicate higher protein expression. Detailed assay validation data, such as detection limits and intra- and inter-assay precision, can be found on the manufacturer’s website (www.olink.com). A total of 275 proteins which passed quality control were assessed with this method.
Cardiac Magnetic Resonance Imaging
CMR scans were conducted using either 1.5-T or 3.0-T scanners available at participating sites manufactured by General Electric, Philips Medical Systems, and Siemens Healthineers. All scans adhered to a standardized protocol defined at the inception of the study and utilized multichannel phased-array chest coils with electrocardiographic gating. Short- and long-axis cine steady-state free precession imaging was performed as previously described (11). T1 mapping in 3 parallel short-axis cuts were carried out using Shortened Modified Look-Locker Inversion recovery (shMOLLI) before and at 3 separate time points after the administration of a gadolinium contrast. The gadolinium contrast was given intravenously as a single dose. Late gadolinium enhancement (LGE) imaging was obtained using a segmented inversion-recovery sequence in both long- and short-axis views. The analysis of cine CMR images was performed at the core laboratory located at Brigham and Women’s Hospital in Boston, Massachusetts, using Medis® (version 7.44, The Netherlands) following guidelines established by the Society for Cardiovascular Magnetic Resonance (Supplementary Methods).
Statistical Analysis
Categorical variables were presented as counts and percentages, while continuous variables were expressed as mean ± standard deviation. Non-normal clinical/imaging variables were log normalized and scaled to mean of 0 and standard deviation of 1. Multivariable linear regression was used to assess the relationship between proteins and clinical/CMR characteristics. All models were adjusted for baseline age, sex, body mass index, hypertension and genotype (presence of MYH7 or MYBPC3 variant). Models were subsequently adjusted for NT-proBNP and BNP. Principal component analyses (PCA) were used for dimensionality reduction for overlapping proteins between left atrial volume index (LAVi) and history of atrial fibrillation. Time to event analyses using Cox proportional hazards regression were conducted for the association of PC 1 and incident atrial fibrillation. The statistical analysis was conducted using commercially available software RStudio 2023. To account for multiple hypothesis testing, a false discovery rate (FDR) of <0.05 was considered statistically significant.
Results
Baseline Characteristics
Among the study participants in the HCMR cohort (n=2,755), 943 had sarcomeric HCM. Resources were available to perform proteomic analyses in in 701 individuals comprising the study cohort. Baseline clinical characteristics for the study cohort are represented in Table 1 and appeared representative of all patients with sarcomeric HCM HCMR (Supplementary Table 1). The mean age was 46.3 +/− 12.0 years, 66% were male, and the majority were white (87%). Hypertension and a history of atrial fibrillation were present in 21% and 11%, respectively. Eighty-four percent carried a pathogenic/likely pathogenic sarcomere variant; 16% had a sarcomere variant of uncertain significance with features favoring pathogenicity. The sarcomere variants were predominantly in MYH7 or MYBPC3 genes.
Table 1:
Baseline characteristics of the study cohort.
| CLINICAL FACTORS | Cohort N = 7011 |
|---|---|
| Age (years) | 46 (12) |
| Male: no. (%) | 463 (66%) |
| Ethnicity (White): no. (%) | 611 (87%) |
| BMI | 27.9 (5.1) |
| Diabetes: no. (%) | 30 (4.3%) |
| Atrial Fibrillation: no. (%) | 77 (11%) |
| Hypertension: no. (%) | 145 (21%) |
| Hyperlipidemia: no. (%) | 121 (17%) |
| Genetic Mutation: no. (%) | |
| MYBPC3 | 377 (54%) |
| MYH7 | 205 (29%) |
| TNNI3 | 31 (4.4%) |
| TNNT2 | 26 (3.7%) |
| MYL2 | 19 (2.7%) |
| TPM1 | 18 (2.6%) |
| MYL3 | 14 (2.0%) |
| ACTC1 | 11 (1.6%) |
| Family History of SCD: no. (%) | 109 (16%) |
| Syncope: no. (%) | 106 (15%) |
| NYHA Class II and above: no. (%) | 197 (29%) |
| SCD: no. (%) | 12 (1.7%) |
| LVOT Gradient > 30 mmHg : no. (%) | 131 (41%) |
| CARDIAC MRI | |
| Maximum LV Wall Thickness (mm): | 21.4 (5.1) |
| Presence of LGE: no. (%) | 478(69%) |
| LGE Burden (% of LV Mass): no. (%) | |
| 15 % | 18 (2.7%) |
| 5–15 % | 104 (15%) |
| <5 % | 556 (82%) |
| LVEF (%) | 64 (9) |
| LV Mass Index (g/m2) | 81 (26) |
| LAVi (ml/m2) | 57 (23) |
mean (SD) for continuous; n (%) for categorical
LVOT: left ventricular outflow tract gradient; LGE: late gadolinium enhancement.
Proteomic associations with cardiac structure and function
To identify circulating proteins associated with cardiac structure and function, we tested the relationship of the 275 proteins measured on the Olink platform (Supplementary Table 2; Mean intra-assay CV across the three Olink proteomic panels was 13%) with core laboratory adjudicated CMR measures of LVEF, LV mass index (LVMi), maximum LV wall thickness and the clinically reported echocardiographic left ventricular outflow tract (LVOT) gradient. Relevant findings are summarized in Figure 1, Table 2 and Supplementary Tables 2–4. After adjustment for age, sex, BMI, HTN and genotype (presence of MYH7 or MYBPC3 variant), both BNP (brain natriuretic peptide) and NT-proBNP (N-terminal pro brain natriuretic peptide) were significantly associated with all four assessed imaging variables, LVEF, LVMi, maximal LV wall thickness and LVOT gradient. Several proteins involved in inflammation and fibrosis, including FGF 23 (fibroblast growth factor 23), CCL15 (c-c motif ligand 15) and MMP2 (matrix metalloproteinase 2) were inversely associated with LVEF and directly associated with LVMi. Pathways associated with vascular remodeling and angiogenesis were also associated with cardiac structure and function. VEGF-D (vascular endothelial growth factor-D) was directly associated with LVOT gradient and LVMi. TM (thrombomodulin) was inversely associated with maximum LV wall thickness and LVMi (Figure 1). All proteomic findings and associations with echo and CMR variables are displayed in Supplementary Tables 3 and 4.
Figure 1: Proteomic associations with cardiac structure, function and myocardial fibrosis.
A) Proteins associated with maximum left ventricular wall thickness, LVMi, LVEF, and LVOT gradient. B) Proteins associated with LGE and ECV. All analyses adjusted for age, sex, BMI, HTN and genotype. FDR was considered significant if <5%. C. Proteins associated with imaging variables after adjustment of BNP.
*P<0.05 after adjustment for clinical variables and BNP
†No proteins remained associated with LVOT obstruction and maximum LV wall thickness after adjustment for BNP
LVMi: left ventricular mass index; LVEF: left ventricular ejection fraction; LVOT: left ventricular outflow tract; LGE: late gadolinium enhancement; ECV: extracellular volume; BMI: body mass index; HTN: hypertension; FDR: false discovery rate
Table 2: Key proteins and their associations with cardiac hypertrophy, fibrosis and hypertrophic cardiomyopathy.
Prior associations with cardiac hypertrophy in small animal models are highlighted: ⬆ or ⬇ denotes evidence from gene knockout model.
| Protein | Biological Function | Cardiac hypertrophy and fibrosis | HCM | Drug Target |
|---|---|---|---|---|
| High Myocardial Expression | ||||
| Brain Natriuretic Peptide and N-terminal pro-brain natriuretic peptide (BNP &NT-proBNP) | Regulation of blood volume and sodium balance | ⬇Cardiac hypertrophy in mice lacking receptor (13) ⬇Fibrosis (14) |
Elevated in HCM, positively associated with wall thickness, LVOT gradient | ✔ |
| Medium Myocardial Expression | ||||
| Angiotensin converting enzyme 2 (ACE2) | Promotes Ang1–7/MasR activation; regulator of blood volume, sodium balance | ⬇Hypertrophy and Fibrosis (15) | Elevated in HCM (16) | ✔ |
|
Thrombomodulin
(TM) |
Inhibits coagulation and inflammation | Upregulated cardiac myocytes with cardiac hypertrophy (17) | -- | ---- |
| Interleukin 18 binding partner (1L18BP) | Inhibits pro-inflammatory cytokine IL18 | ⬇Hypertrophy in adrenergic induced model (18) | -- | ✔ |
|
Tenascin
(TNC) |
Inflammation, immunity and angiogenesis | ⬆Fibrosis in pressure overload model (19) | Correlated with HF events in HCM (20) | ✔ |
| Adrenomedullin (ADM) | Cardiac: antagonizes effects of ANGII Vascular: maintains endothelial barrier |
⬇Hypertrophy and fibrosis in pressure overload model (21) | Increased in HCM vs. controls (22) | ✔ |
| Matrix metalloproteinase 2 (MMP2) | Extracellular matrix protein, cleaves collagen | ⬇ ⬆Hypertrophy in pressure overload model (23) | Elevated in HCM, LAV, LVEF (24) | ✔ |
|
Spondin 1
(SPON1) |
Cell adhesion and extracellular matrix composition | ---- | -- | ---- |
| Tumor necrosis factor receptor 2 (TNFR2) | Binds TNF-alpha; anti-inflammatory | ⬇Hypertrophy(25) (26) and Fibrosis | -- | ---- |
| EPH Receptor 4 (EPHB4) | Regulates vascular development | -- | -- | |
| Vascular endothelial growth factor-D (VEGFD) | Angiogenesis | VEGF receptor decoy reduces cardiac hypertrophy and transitions to LV dilation (27) | Correlated with HCM clinical profiles (28) | ---- |
| Low Myocardial Expression | ||||
| Chitinase 3-like protein 1 (CHI3L1) | Inflammation and tissue remodeling | -- | -- | ✔ |
| Fibroblast Growth Factor 23 (FGF23) | Phosphate and vitamin D metabolism | Induces LVH (29) | -- | ✔ |
| Stem cell factor (SCF) | Cytokine in hematopoiesis | ⬇Hypertrophy and fibrosis in MI (30) | -- | ✔ |
| Interleukin 1 receptor like 2 (IL1RL2) | Modulates immune response | --- | -- | ✔ |
| Not expressed in Myocardial Tissue | ||||
| Integrin subunit beta 2 (ITGB2) | Leukocyte adhesion and migration | Promotes cardiac hypertrophy (31) | -- | ✔ |
| Triggering receptor expressed on myeloid like cells 2 (TREML2/TLT2) | Innate and adaptive immune response | -- | -- | ---- |
| Chemokine ligand 15 (CCL15) | Chemotaxis of T cells and monocytes | -- | -- | ✔ |
| Alpha 1 microglobulin/bikunin precursor (AMBP) | Anti-oxidant; free radical scavenger; tissue repair and inflammation | -- | -- | ✔ |
HCM: hypertrophic cardiomyopathy; AF: atrial fibrillation; LVOT: left ventricular outflow tract gradient
Proteomic associations with cardiac fibrosis
CMR allows non-invasive characterization of myocardial tissue characteristics, including assessment of replacement fibrosis by analyzing late gadolinium enhancement (LGE) and interstitial fibrosis by analyzing extracellular volume fraction (ECV). LGE was directly associated with BNP, NT-proBNP, ACE2 (angiotensin converting enzyme 2), a neurohormonal modulator of cardiac hypertrophy and fibrosis (Figure 1). Several proteins were inversely associated with LGE including TNFR2 (tumor necrosis factor receptor 2), a modulator of inflammation through its effects of TNF-alpha (tumor necrosis factor-alpha). ECV was associated with BNP, NT-proBNP and several inflammatory proteins including SCF (Stem-Cell Factor) a cytokine and chemoattractant that was also inversely associated with SCD risk score (below), and ITGB2 (integrin subunit beta-2), a facilitator of leukocyte migration. As seen when assessing measures of cardiac structure and function, we found associations between fibrosis and proteins central to endothelial biology including ADM (adrenomedullin), a potent vasodilator and VEGF-D.
Because biological pathways may be shared by different aspects of adverse cardiac remodeling, we investigated whether protein signals were significantly associated with single or multiple features. As summarized in Figure 2, BNP and NT-proBNP were the only proteins associated with all six imaging variables, spanning structure, function, and fibrosis. In addition, we identified novel proteins associated with multiple traits included ACE2 and TM (LGE, LVMi, maximum wall thickness); VEGF-D (LVOT gradient, LVMi, LGE).
Figure 2: Overlapping proteomic associations with cardiac structure, function and myocardial fibrosis.
Upset plot showing shared protein findings between cardiac imaging variables.
LVOT: left ventricular outflow gradient; LGE: late gadolinium enhancement; ECV: extracellular volume; LVMi: left ventricular mass index.
Proteomic associations with ESC risk score for Sudden Cardiac Death.
The European Society of Cardiology (ESC) risk score for SCD in HCM uses several clinical variables to estimate the 5-year probability of SCD. Components include current age, left ventricular outflow (LVOT) gradient, left atrial (LA) diameter, maximum left ventricular wall thickness, non-sustained ventricular tachycardia (NSVT), prior syncope, and family history of SCD (32). Of the 701 patients with proteomic data, 259 individuals had sufficient data to calculate an ESC risk score (other participants were typically missing LA diameter and/or maximal wall thickness data from clinical echocardiograms used in HCMR). Mean ESC risk score in this subgroup was 2.6 ± 0.6 which is considered to be low risk for SCD (primary prevention implantable cardioverter defibrillator [ICD] not recommended). Of the 275 proteins assayed, six were significantly associated with the ESC risk score (Figure 3A; Supplementary Table 5). BNP and NT-proBNP were directly associated with ESC risk score. This association was likely driven by positive correlations with maximum LV wall thickness and left atrial diameter (Figure 4B). Four proteins were inversely associated with ESC risk score, DLK-1 (non-canonical delta NOTCH ligand), SCF (stem cell factor), IL1RT1 (interleukin 1-receptor type 1), IGFBP-1 (insulin like growth factor binding protein 1), with increasing protein levels associated with lower ESC risk score. DLK-1was positively correlated with age and negatively correlated with maximum wall thickness and left atrial diameter; SCF was negatively correlated with LVOT gradient. As previously mentioned, SCF was also inversely associated with ECV, a marker of interstitial fibrosis.
Figure 3: Proteomic associations with ESC Sudden Cardiac Death Risk Score.
A. Proteins associated with ESC risk score. Effect size (beta) represents change in ESC risk score per log2 increase in protein value. B. Correlations of proteins with individual components of ESC risk score (Pearson coefficient for continuous variables and point-biserial coefficient for dichotomous variables).
*FDR < 5% was used to account for multiple hypothesis testing for all comparisons.
ESC: European Society Congress; NSVT: Non-sustained ventricular tachycardia; LVOT: left ventricular outflow tract
Figure 4: Overlapping proteomic associations between left atrial volume index and prevalent atrial fibrillation.
Twenty-six proteins were associated with prevalent AF and 4 proteins were associated with LAVI only. Fifteen proteins are significantly associated with both LAVi and prevalent atrial fibrillation at baseline (FDR<5% for both). Twelve of these proteins are directly associated with LAVi and AF and 3 are inversely associated with LAVi and AF. Models adjusted for age, sex, BMI, HTN and genotype status.
LAVi: left atrial volume index; AF: atrial fibrillation; BMI: body mass index
Protein Associations with Left Atrial Remodeling and Atrial Fibrillation
Individuals with HCM are at heightened risk for both AF and associated cardioembolic events. Left atrial enlargement has been associated with increased risk of AF (33). Therefore, we postulated that identifying proteins associated with prevalent AF and left atrial remodeling, as reflected by left atrial volume indexed to BSA (LAVi) derived from CMR, may help identify predictors of risk of incident AF (Supplementary Figure 1). After multivariable adjustment, 15 proteins were significantly associated with both LAVi and prevalent AF (Figure 4; Supplementary Table 6), including markers of atrial stretch (BNP, NT-proBNP); fibrosis (FGF-23), and regulators of vascular function (NOTCH-3). While some of these proteins were also associated with LV structure and function on CMR, other proteins were uniquely associated with LAVi and AF, including IGFB2 (insulin like growth factor binding protein 2), a regulator of adipogenesis and insulin sensitivity.
Principal component analysis can help reduce complexity of high dimensional data to summarize and retain the most important features of a dataset into individual components or factors. Applying PCA to the 15 proteins associated with both LAVi and prevalent AF, we found that the top PC (PC1) explained 30% of the variation of these proteins (Figure 5). The proteins which contribute the most to PC1 are MMP2, SPON1 and NOTCH-3. To assess proteins associated with incident AF, we performed time to event analyses using Cox-proportional hazards model. There were 27 incident AF events with mean time of follow up of 6 years. One standard deviation increase in PC1 was associated with an increased risk of incident AF (HR 1.65; p = 0.003) in unadjusted analyses. After adjustment for covariates, PC1 remained associated with incident AF (HR 1.53; p value 0.02). Individual associations between proteins and incident AF are shown in Supplementary Table 7.
Figure 5: Proteomic associations with incident atrial fibrillation.
A. Principal component analyses of the 15 proteins overlapping LAVi and prevalent AF. PC1 accounted for 29.8% of the variation of the 15 protein levels. B. Proteins and PC1 loadings with the top weighted proteins: MMP2, SPON1, NOTCH3. C. Time to event analyses for PC1 and incident AF in: 1) unadjusted and 2) adjusted model (age, sex, BMI, HTN and genotype). Cox regression models were used to assess risk of incident AF per standard deviation increase in PC1.
PC: Principal component; AF: atrial fibrillation; BMI: body mass index; HTN: hypertension
Discussion
In this study of patients with sarcomeric HCM, we describe plasma proteomic signatures associated with adverse cardiac remodeling, myocardial fibrosis, estimated SCD risk, and prevalent and incident AF. Some of the proteins identified are involved in pathways of known relevance to HCM, including hemodynamic stress, neurohormonal modulation, and fibrosis. Other proteins are involved in pathways less established in HCM and possibly more peripheral to the cardiomyocyte, such as vascular biology, endothelial function, and inflammation. Prior proteomic investigations in HCM have largely leveraged myocardial tissue samples from myectomy performed in highly symptomatic patients in advanced stages of disease. These tissue studies revealed alterations in myocardial structural proteins, calcium signaling and metabolism pathways among others (4,34) but are difficult to apply to clinical practice due to lack of routine accessibility to myocardial tissue. Recently, Shimada et al. used a broad scale plasma proteomic platform to identify circulating proteomic features of HCM, highlighting upregulated intracellular signaling pathways in HCM including Ras/MAPK (mitogen-activated protein kinase) and transforming growth factor beta compared to patients with hypertensive LVH. Their cohort was older than this study (mean age 61 vs 46 years, respectively), and likely more heterogeneous as only a small subset underwent genotyping (9). By contrast, this study leveraged a large cohort of genotyped individuals with sarcomeric HCM and identified several proteins which had not been previously recognized in the context of HCM or cardiac hypertrophy. By highlighting additional pathways and potential mechanisms associated with key clinical features of HCM, our findings provide support for the use of protein profiling to identify biomarkers of remodeling.
BNP and NT-proBNP provided the most consistent signal and were identified in all investigated protein-disease associations, from cardiac structure/function, SCD risk and the development of atrial fibrillation. BNP and NT-proBNP are induced by increased myocardial wall stress, elevated ventricular pressures, and atrial stretch. They are well-established prognostic biomarkers in heart failure and have previously been studied in HCM (35). Despite assaying several hundred proteins pertinent to cardiovascular disease, these well-established biomarkers performed the best across all clinical and imaging markers, highlighting their relevance in HCM) and helping to confirm the validity of the Olink proteomics platform.
The renin-angiotensin-aldosterone (RAAS) pathway plays a critical role in cardiac hypertrophy and fibrosis and in this study soluble ACE2 was associated with LVMi and LGE. ACE2 catalyzes the conversion of angiotensin II to angiotensin [1–7], a vasoactive peptide which exerts its effects via the Mas receptor pathway. This enzymatic step helps counterbalance the vasoconstrictive effect of angiotensin II and stimulation of renin-aldosterone pathway. In a transcriptomic analysis of septal myectomy tissue samples compared to healthy heart donors, ACE2 was seen to be most upregulated gene in HCM (16). Collectively these findings build evidence that ACE2 may be a modifier of HCM severity and a biomarker of disease progression. In addition, we found that adrenomedullin (ADM), a peptide hormone with implicated roles in regulating neurohormonal activation in heart failure, was associated ECV, a marker of cardiac fibrosis. ADM is a potent vasodilator and promotes natriuresis through its actions on the vasculature and kidney and counteracts the sodium retaining effects of the RAAS pathway (36). These findings highlight the role of proteins in the neurohormonal pathways as potential key drivers in cardiac remodeling in HCM.
To help contextualize our results, we reviewed the literature for the other proteins found to associate with cardiac remodeling, fibrosis, atrial fibrillation, and SCD risk. (Table 2; Figure 6). These proteins could be broadly grouped to impact endothelial/vascular function, inflammation/immunity, and fibrosis. Coronary microvascular dysfunction is well-described in HCM (37) and endothelial dysfunction may also be present (38). These pathological processes can exacerbate symptoms and the risk of heart failure. We found that several proteins involved in maintaining vascular and endothelial health, notably thrombomodulin (TM) and vascular endothelial growth factor-D (VEGF-D), were associated with adverse remodeling in HCM. Both proteins were associated with LVH and myocardial fibrosis and VEGF -D was additionally associated with LVOT obstruction and AF. TM is localized to the endothelium and plays a prominent role in regulating coagulation, cell trafficking, inflammation and angiogenesis (39) VEGF-D also regulates blood vessel remodeling and angiogenesis. Angiogenesis may result from cardiac hypertrophy and itself has been shown to promote hypertrophy of cardiac myocytes in preclinical models (40). While the imbalance of angiogenesis and cardiac hypertrophy may contribute to decompensated heart failure (41), the degree to which these vascular processes are compensatory or stem directly from HCM-related myocardial remodeling is unknown and requires further study.
Figure 6: Main proteins and pathways associated with clinical and imaging features in sarcomeric HCM.
ESC: European society of cardiology; SCD: sudden cardiac death
Individuals with HCM have an increased risk of SCD, although underlying mechanisms are not fully characterized. High risk cardiac MRI features and clinical risk scores can help risk stratify individuals with HCM for SCD. LGE on cardiac MRI represents areas of replacement myocardial fibrosis that may serve as a nidus for ventricular arrythmias. The presence of extensive LGE is considered a potential modifier for stratifying SCD risk in the ACC/AHA guidelines (42). In addition to proteins that were also associated with cardiac structure, such as natriuretic proteins and ACE2, we found that several unique proteins, including TNFR2 and IL18BP, were inversely associated with LGE. TNFR2 is a regulator of TNF, a critical protein in inflammatory responses. TNFR2 has been shown to activate transmembrane TNF-α, which may promote anti-inflammatory and anti-hypertrophic effects, counter to soluble TNF-α (25). Similarly, IL18BP is an inhibitor of IL18, a pro-inflammatory cytokine shown to promote pathologic cardiac remodeling in mouse model of cardiac hypertrophy (43).
In addition to protein associations with LGE, we found several associations with the 5-year ESC risk score for SCD in HCM. Stem cell factor (SCF) was inversely associated with the ESC risk score, driven by its inverse associations with LVOT gradient and LVMi. Moreover, SCF was inversely associated with ECV on CMR, a marker of diffuse interstitial fibrosis. SCF is a cytokine which exists in both transmembrane and soluble forms. It helps regulate hematopoiesis (12) and functions as a chemoattractant for mast cells and monocytes. In addition, SCF is expressed in fibroblasts and endothelial cells and binds to its regulatory partner C-KIT, which has previously seen to be differentially expressed in HCM vs healthy controls (44). TNFR2, IL18BP and SCF were all inversely associated with BNP and NT-proBNP. However, their association with SCD risk markers remained significant even after adjustment for the natriuretic proteins, suggesting they are orthogonal biomarkers and potentially represent novel pathways in the development of cardiac fibrosis and SCD risk in HCM.
AF is responsible for substantial morbidity and mortality in HCM, due to higher prevalence and higher risk of stroke irrespective of traditional risk stratification tools, such as CHADS2Vasc score (45). Furthermore, treatments for AF may be less effective in HCM due to the driving influence of the underlying cardiomyopathy and more advanced atrial remodeling (46). We found that several proteins associated with both LAVi and prevalent AF were also associated with LV remodeling, including NT-proBNP, ACE2, VEGF-D and MMP2, However, we also found proteins that were associated only with AF and not associated with left ventricular remodeling. IGFBP2 is involved in glucose and lipid metabolism. As such, its association with AF may be mediated through its role in cardiometabolic health.
Proteins associated with future development of AF may provide additional insight into disease progression in HCM. We identified proteins involved in inflammatory and fibrotic processes, such as MMP2 and SPON1, to be associated with incident AF. MMP2 is a matrix metalloproteinase, implicated in collagen turnover and fibrosis. MMP2 has also been shown to be increased in individuals with HCM versus controls (47), and elevated MMP2 has been associated with longer durations of AF as well recurrent/persistent AF in individuals who underwent cardioversion (48). SPON1 is a regulator of macrophage proliferation. Additionally, SPON1 has been associated AF in other populations and by other modalities. For example, SPON1 levels have been associated with incident AF in the elderly (49), and genetic polymorphisms in SPON1 were associated with AF in a genome wide association study (50). Our findings implicate MMP2 and SPON1 as potential markers for the development of AF in HCM and provide more evidence that these proteins may be involved in HCM pathophysiology.
Limitations of the Study
Proteomic profiling was performed with established Olink panels that are enriched for biological processes associated with cardiovascular disease, limiting the number of proteins and pathways detected. This targeted approach may bias findings. Further, Cardiac Troponin T (cTnT), a an important biomarker in HCM, was not assayed in the current study. However, cTnT was previously measured in the HCMR cohort and shown to be associated with cardiac MRI variables such as LGE and ECV (10). We studied sarcomeric HCM and findings may not be generalizable to non-sarcomeric HCM. However, to our knowledge, our study is the largest plasma proteomics study in sarcomeric HCM, and may help identify more direct associations with HCM-specific pathology. Since we are measuring proteins in plasma, the origin of these proteins, whether reflecting cardiac or peripheral biology secondary to HCM, is not known. Finally, additional proteomic studies in this population are needed to validate findings and to determine whether measurement of circulating proteins add prognostic value for long term outcomes above clinical risk factors and well-established cardiac biomarkers such as BNP and NT-proBNP.
Conclusions
Proteomic profiling of sarcomeric HCM identified proteins associated with adverse clinical and imaging features. These proteins are part of both established pathways in HCM, including natriuretic peptides, neurohormonal activation and fibrosis, and less studied pathways, including vascular biology, endothelial function and inflammation. These findings highlight the potential for plasma profiling to identify biomarkers of risk and to gain further insights into the pathophysiology of HCM. These insights may lead to the development of novel, pathway-based therapies. Further studies are needed to assess these associations in non-sarcomeric HCM and to identify proteins associated with long term outcomes.
Supplementary Material
Clinical Perspectives.
What’s new?
BNP and NT-proBNP are robustly associated with cardiac remodeling and clinical outcomes in sarcomeric HCM.
Additional biomarkers in pathways involving neurohormonal activation, inflammation and fibrosis were associated with abnormal morphology and adverse clinical features, including the development of atrial fibrillation and increased SCD risk score.
Clinical implications:
Identifying circulating proteomic signatures associated with adverse clinical and imaging phenotypes may aid in risk stratification in HCM.
Proteins implicated in inflammation and endothelial function represent alternative pathways in cardiac remodeling and fibrosis, providing opportunities to identify novel therapeutic targets in HCM.
Acknowledgements:
UT acknowledges support from the NHLBI mentored career development award: K08 HL161445-01A1. SN acknowledges support from the Oxford NIHR Biomedical Research Centre and from the British Heart Foundation Centre of Research Excellence. HW acknowledges support from the British Heart Foundation Centre of Research Excellence and from CureHeart, the British Heart Foundation’s Big Beat Challenge award (BBC/F/21/220106)
Sources of Funding
National Institutes of Health, National Heart Lung Blood Institute U01HL117006-01A1, Cytokinetics, British Heart Foundation, Center of Research Excellence, and Oxford NIHR Biomedical Research Centre
Disclaimer
The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the National Heart, Lung, and Blood Institute; the National Institutes of Health; or the U.S. Department of Health and Human Services.
Non-standard Abbreviations and Acronyms
- HCM
Hypertrophic Cardiomyopathy
- AF
Atrial Fibrillation
- HF
Heart Failure
- SCD
Sudden Cardiac Death
- CMR
Cardiac Magnetic Resonance
- LGE
Late Gadolinium Enhancement
- LVMi
Left Ventricular Mass Index
- LVOT
Left Ventricular Outflow Tract
- ESC
European Society of Cardiology
- PCA
Principal Component Analysis
- LAVi
Left Atrial Volume Indexed to Body Surface Area
- BMI
Body Mass Index
- PEA
Proximity Extension Assay
- NPX
Normalized Protein expression
Footnotes
Disclosure Statement:
The authors have nothing to disclose.
References
- 1.Maron BJ, Gardin JM, Flack JM, Gidding SS, Kurosaki TT, Bild DE. Prevalence of hypertrophic cardiomyopathy in a general population of young adults. Echocardiographic analysis of 4111 subjects in the CARDIA Study. Coronary Artery Risk Development in (Young) Adults. Circulation. 1995. Aug 15;92(4):785–9. [DOI] [PubMed] [Google Scholar]
- 2.Nurmohamed NS, Belo Pereira JP, Hoogeveen RM, Kroon J, Kraaijenhof JM, Waissi F, et al. Targeted proteomics improves cardiovascular risk prediction in secondary prevention. Eur Heart J. 2022. Apr 21;43(16):1569–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ganz P, Heidecker B, Hveem K, Jonasson C, Kato S, Segal MR, et al. Development and Validation of a Protein-Based Risk Score for Cardiovascular Outcomes Among Patients With Stable Coronary Heart Disease. JAMA. 2016. Jun 21;315(23):2532–41. [DOI] [PubMed] [Google Scholar]
- 4.Coats CJ, Heywood WE, Virasami A, Ashrafi N, Syrris P, Dos Remedios C, et al. Proteomic Analysis of the Myocardium in Hypertrophic Obstructive Cardiomyopathy. Circ Genomic Precis Med. 2018. Dec;11(12):e001974. [DOI] [PubMed] [Google Scholar]
- 5.Schuldt M, Pei J, Harakalova M, Dorsch LM, Schlossarek S, Mokry M, et al. Proteomic and Functional Studies Reveal Detyrosinated Tubulin as Treatment Target in Sarcomere Mutation-Induced Hypertrophic Cardiomyopathy. Circ Heart Fail. 2021. Jan;14(1):e007022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Tucholski T, Cai W, Gregorich ZR, Bayne EF, Mitchell SD, McIlwain SJ, et al. Distinct hypertrophic cardiomyopathy genotypes result in convergent sarcomeric proteoform profiles revealed by top-down proteomics. Proc Natl Acad Sci U S A. 2020. Oct 6;117(40):24691–700. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Katz DH, Tahir UA, Ngo D, Benson MD, Gao Y, Shi X, et al. Multiomic Profiling in Black and White Populations Reveals Novel Candidate Pathways in Left Ventricular Hypertrophy and Incident Heart Failure Specific to Black Adults. Circ Genomic Precis Med. 2021. Jun;14(3):e003191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ngo D, Sinha S, Shen D, Kuhn EW, Keyes MJ, Shi X, et al. Aptamer-Based Proteomic Profiling Reveals Novel Candidate Biomarkers and Pathways in Cardiovascular Disease. Circulation. 2016. Jul 26;134(4):270–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Shimada YJ, Raita Y, Liang LW, Maurer MS, Hasegawa K, Fifer MA, et al. Comprehensive Proteomics Profiling Reveals Circulating Biomarkers of Hypertrophic Cardiomyopathy. Circ Heart Fail. 2021. Jul;14(7):e007849. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Neubauer S, Kolm P, Ho CY, Kwong RY, Desai MY, Dolman SF, et al. Distinct Subgroups in Hypertrophic Cardiomyopathy in the NHLBI HCM Registry. J Am Coll Cardiol. 2019. Nov 12;74(19):2333–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kramer CM, Appelbaum E, Desai MY, Desvigne-Nickens P, DiMarco JP, Friedrich MG, et al. Hypertrophic Cardiomyopathy Registry: The rationale and design of an international, observational study of hypertrophic cardiomyopathy. Am Heart J. 2015. Aug;170(2):223–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Assarsson E, Lundberg M, Holmquist G, Björkesten J, Thorsen SB, Ekman D, et al. Homogenous 96-Plex PEA Immunoassay Exhibiting High Sensitivity, Specificity, and Excellent Scalability. PLOS ONE. 2014. Apr 22;9(4):e95192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Oliver PM, Fox JE, Kim R, Rockman HA, Kim HS, Reddick RL, et al. Hypertension, cardiac hypertrophy, and sudden death in mice lacking natriuretic peptide receptor A. Proc Natl Acad Sci U S A. 1997. Dec 23;94(26):14730–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Tamura N, Ogawa Y, Chusho H, Nakamura K, Nakao K, Suda M, et al. Cardiac fibrosis in mice lacking brain natriuretic peptide. Proc Natl Acad Sci U S A. 2000. Apr 11;97(8):4239–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Moritani T, Iwai M, Kanno H, Nakaoka H, Iwanami J, Higaki T, et al. ACE2 deficiency induced perivascular fibrosis and cardiac hypertrophy during postnatal development in mice. J Am Soc Hypertens JASH. 2013;7(4):259–66. [DOI] [PubMed] [Google Scholar]
- 16.Bos JM, Hebl VB, Oberg AL, Sun Z, Herman DS, Teekakirikul P, et al. Marked Up-Regulation of ACE2 in Hearts of Patients With Obstructive Hypertrophic Cardiomyopathy: Implications for SARS-CoV-2-Mediated COVID-19. Mayo Clin Proc. 2020. Jul;95(7):1354–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Li YH, Chung HC, Luo CY, Chao TH, Shyu KG, Shi GY, et al. Thrombomodulin is upregulated in cardiomyocytes during cardiac hypertrophy and prevents the progression of contractile dysfunction. J Card Fail. 2010. Dec;16(12):980–90. [DOI] [PubMed] [Google Scholar]
- 18.Murray DR, Mummidi S, Valente AJ, Yoshida T, Somanna NK, Delafontaine P, et al. β2 adrenergic activation induces the expression of IL-18 binding protein, a potent inhibitor of isoproterenol induced cardiomyocyte hypertrophy in vitro and myocardial hypertrophy in vivo. J Mol Cell Cardiol. 2012. Jan 1;52(1):206–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Song L, Wang L, Li F, Yukht A, Qin M, Ruther H, et al. Bone Marrow-Derived Tenascin-C Attenuates Cardiac Hypertrophy by Controlling Inflammation. J Am Coll Cardiol. 2017. Sep 26;70(13):1601–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kitaoka H, Kubo T, Baba Y, Yamasaki N, Matsumura Y, Furuno T, et al. Serum tenascin-C levels as a prognostic biomarker of heart failure events in patients with hypertrophic cardiomyopathy. J Cardiol. 2012. Mar;59(2):209–14. [DOI] [PubMed] [Google Scholar]
- 21.Niu P, Shindo T, Iwata H, Iimuro S, Takeda N, Zhang Y, et al. Protective effects of endogenous adrenomedullin on cardiac hypertrophy, fibrosis, and renal damage. Circulation. 2004. Apr 13;109(14):1789–94. [DOI] [PubMed] [Google Scholar]
- 22.Hamada M, Shigematsu Y, Kawakami H, Minamino N, Kangawa K, Matsuo H, et al. Increased plasma levels of adrenomedullin in patients with hypertrophic cardiomyopathy: its relation to endothelin-I, natriuretic peptides and noradrenaline. Clin Sci Lond Engl 1979. 1998. Jan;94(1):21–8. [DOI] [PubMed] [Google Scholar]
- 23.Matsusaka H, Ide T, Matsushima S, Ikeuchi M, Kubota T, Sunagawa K, et al. Targeted Deletion of Matrix Metalloproteinase 2 Ameliorates Myocardial Remodeling in Mice With Chronic Pressure Overload. Hypertension. 2006. Apr;47(4):711–7. [DOI] [PubMed] [Google Scholar]
- 24.Lombardi R, Betocchi S, Losi MA, Tocchetti CG, Aversa M, Miranda M, et al. Myocardial collagen turnover in hypertrophic cardiomyopathy. Circulation. 2003. Sep 23;108(12):1455–60. [DOI] [PubMed] [Google Scholar]
- 25.Miao K, Zhou L, Ba H, Li C, Gu H, Yin B, et al. Transmembrane tumor necrosis factor alpha attenuates pressure-overload cardiac hypertrophy via tumor necrosis factor receptor 2. PLoS Biol. 2020. Dec;18(12):e3000967. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Duerrschmid C, Crawford JR, Reineke E, Taffet GE, Trial J, Entman ML, et al. TNF Receptor 1 Signaling is Critically Involved in Mediating Angiotensin-II-induced Cardiac Fibrosis. J Mol Cell Cardiol. 2013. Apr;57:59–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Izumiya Y, Shiojima I, Sato K, Sawyer DB, Colucci WS, Walsh K. Vascular Endothelial Growth Factor Blockade Promotes the Transition From Compensatory Cardiac Hypertrophy to Failure in Response to Pressure Overload. Hypertension. 2006. May;47(5):887–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Pudil R, Vasatova M, Fucikova A, Rehulkova H, Rehulka P, Palicka V, et al. Vascular Endothelial Growth Factor Is Associated with the Morphologic and Functional Parameters in Patients with Hypertrophic Cardiomyopathy. BioMed Res Int. 2015;2015:762950. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Faul C, Amaral AP, Oskouei B, Hu MC, Sloan A, Isakova T, et al. FGF23 induces left ventricular hypertrophy. J Clin Invest. 2011. Nov 1;121(11):4393–408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Xiang FL, Lu X, Hammoud L, Zhu P, Chidiac P, Robbins J, et al. Cardiomyocyte-Specific Overexpression of Human Stem Cell Factor Improves Cardiac Function and Survival After Myocardial Infarction in Mice. Circulation. 2009. Sep 22;120(12):1065–74. [DOI] [PubMed] [Google Scholar]
- 31.Liu L, Zhao Q, Kong M, Mao L, Yang Y, Xu Y. Myocardin-related transcription factor A regulates integrin beta 2 transcription to promote macrophage infiltration and cardiac hypertrophy in mice. Cardiovasc Res. 2022. Feb 21;118(3):844–58. [DOI] [PubMed] [Google Scholar]
- 32.O’Mahony C, Jichi F, Pavlou M, Monserrat L, Anastasakis A, Rapezzi C, et al. A novel clinical risk prediction model for sudden cardiac death in hypertrophic cardiomyopathy (HCM Risk-SCD). Eur Heart J. 2014. Aug 7;35(30):2010–20. [DOI] [PubMed] [Google Scholar]
- 33.Kramer CM, DiMarco JP, Kolm P, Ho CY, Desai MY, Kwong RY, et al. Predictors of Major Atrial Fibrillation Endpoints in the National Heart, Lung, and Blood Institute HCMR. JACC Clin Electrophysiol. 2021. Nov;7(11):1376–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Moore J, Ewoldt J, Venturini G, Pereira AC, Padilha K, Lawton M, et al. Multi-Omics Profiling of Hypertrophic Cardiomyopathy Reveals Altered Mechanisms in Mitochondrial Dynamics and Excitation–Contraction Coupling. Int J Mol Sci. 2023. Jan;24(5):4724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Coats CJ, Gallagher MJ, Foley M, O’Mahony C, Critoph C, Gimeno J, et al. Relation between serum N-terminal pro-brain natriuretic peptide and prognosis in patients with hypertrophic cardiomyopathy. Eur Heart J. 2013. Aug 21;34(32):2529–37. [DOI] [PubMed] [Google Scholar]
- 36.Nishikimi T, Yoshihara F, Horinaka S, Kobayashi N, Mori Y, Tadokoro K, et al. Chronic administration of adrenomedullin attenuates transition from left ventricular hypertrophy to heart failure in rats. Hypertens Dallas Tex 1979. 2003. Nov;42(5):1034–41. [DOI] [PubMed] [Google Scholar]
- 37.Petersen SE, Jerosch-Herold M, Hudsmith LE, Robson MD, Francis JM, Doll HA, et al. Evidence for Microvascular Dysfunction in Hypertrophic Cardiomyopathy. Circulation. 2007. May 8;115(18):2418–25. [DOI] [PubMed] [Google Scholar]
- 38.Heffernan KS, Napolitano CA, Maron MS, Patvardhan EA, Patel AR, Pandian NG, et al. Peripheral vascular endothelial function in patients with hypertrophic cardiomyopathy. Am J Cardiol. 2010. Jan 1;105(1):112–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Martin FA, Murphy RP, Cummins PM. Thrombomodulin and the vascular endothelium: insights into functional, regulatory, and therapeutic aspects. Am J Physiol - Heart Circ Physiol. 2013. Jun 15;304(12):H1585–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Walsh K, Shiojima I. Cardiac growth and angiogenesis coordinated by intertissue interactions. J Clin Invest. 2007. Nov 1;117(11):3176–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Shiojima I, Sato K, Izumiya Y, Schiekofer S, Ito M, Liao R, et al. Disruption of coordinated cardiac hypertrophy and angiogenesis contributes to the transition to heart failure. J Clin Invest. 2005. Aug;115(8):2108–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Ommen SR, Mital S, Burke MA, Day SM, Deswal A, Elliott P, et al. 2020 AHA/ACC Guideline for the Diagnosis and Treatment of Patients With Hypertrophic Cardiomyopathy: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2020. Dec 22;142(25):e558–631. [DOI] [PubMed] [Google Scholar]
- 43.Xiao H, Li H, Wang JJ, Zhang JS, Shen J, An XB, et al. IL-18 cleavage triggers cardiac inflammation and fibrosis upon β-adrenergic insult. Eur Heart J. 2018. Jan 1;39(1):60–9. [DOI] [PubMed] [Google Scholar]
- 44.Sonnenschein K, Fiedler J, de Gonzalo-Calvo D, Xiao K, Pfanne A, Just A, et al. Blood-based protein profiling identifies serum protein c-KIT as a novel biomarker for hypertrophic cardiomyopathy. Sci Rep. 2021. Jan 19;11(1):1755. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Olivotto I, Cecchi F, Casey SA, Dolara A, Traverse JH, Maron BJ. Impact of atrial fibrillation on the clinical course of hypertrophic cardiomyopathy. Circulation. 2001. Nov 20;104(21):2517–24. [DOI] [PubMed] [Google Scholar]
- 46.Tsang TS, Barnes ME, Bailey KR, Leibson CL, Montgomery SC, Takemoto Y, et al. Left atrial volume: important risk marker of incident atrial fibrillation in 1655 older men and women. Mayo Clin Proc. 2001. May;76(5):467–75. [DOI] [PubMed] [Google Scholar]
- 47.Noji Y, Shimizu M, Ino H, Higashikata T, Yamaguchi M, Nohara A, et al. Increased circulating matrix metalloproteinase-2 in patients with hypertrophic cardiomyopathy with systolic dysfunction. Circ J Off J Jpn Circ Soc. 2004. Apr;68(4):355–60. [DOI] [PubMed] [Google Scholar]
- 48.Kato K, Fujimaki T, Yoshida T, Oguri M, Yajima K, Hibino T, et al. Impact of matrix metalloproteinase-2 levels on long-term outcome following pharmacological or electrical cardioversion in patients with atrial fibrillation. EP Eur. 2009. Mar 1;11(3):332–7. [DOI] [PubMed] [Google Scholar]
- 49.Norby FL, Tang W, Pankow JS, Lutsey PL, Alonso A, Steffan B, et al. Proteomics and Risk of Atrial Fibrillation in Older Adults (From the Atherosclerosis Risk in Communities [ARIC] Study). Am J Cardiol. 2021. Dec 15;161:42–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Miyazawa K, Ito K, Ito M, Zou Z, Kubota M, Nomura S, et al. Cross-ancestry genome-wide analysis of atrial fibrillation unveils disease biology and enables cardioembolic risk prediction. Nat Genet. 2023. Feb;55(2):187–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.