Review in Translational Cardiology: MicroRNAs and Myocardial Fibrosis in Aortic Valve Stenosis, a Deep Insight on Left Ventricular Remodeling.

MicroRNAs (miRNAs) are a huge class of noncoding RNAs that regulate protein-encoding genes (degradation/inhibition of translation). miRNAs are nowadays recognized as regulators of biological processes underneath cardiovascular disorders including hypertrophy, ischemia, arrhythmias, and valvular disease. In particular, circulating miRNAs are promising biomarkers of pathology. This review gives an overview of studies in aortic valve stenosis (AS), exclusively considering myocardial remodeling processes. We searched through literature (till September 2016), all studies and reviews involving miRNAs and AS (myocardial compartment). Although at the beginning of a new era, clear evidences exist on the potential diagnostic and prognostic implementation of miRNAs in the clinical setting. In particular, for AS, miRNAs are modulators of myocardial remodeling and hypertrophy. In our experience, here presented in summary, the principal findings of our research were a confirm of the pathophysiological role in AS of miRNA-21, in particular, the interdependence between textural miRNA-21 and fibrogenic stimulus induced by an abnormal left ventricular pressure overload. Moreover, circulating miRNA-21 (biomarker) levels are able to reflect the presence of significant myocardial fibrosis (MF). Thus, the combined evaluation of miRNA-21, a marker of MF, and hypertrophy, together with advanced echocardiographic imaging (two-dimensional speckle tracking), could fulfill many existing gaps, renewing older guidelines paradigms, also allowing a better risk prognostic and diagnostic strategies.

INTRODUCTION

Valvular heart disease (VHD) is a major cause of cardiovascular mortality. Degenerative aortic valve stenosis (AS) has become the most common cause of VHD in the Western world. The prevalence of moderate or severe AS in patients aged 75–95 years equals 2.8%.[123]

AORTIC VALVE DISEASE PATHOPHYSIOLOGY: LEFT VENTRICULAR REMODELLING

The pathophysiological mechanisms of valve calcification are the result of an active and heterogeneous process. Basement membrane disruption, chronic inflammation, lipid deposition, osteoblastic transition of valve interstitial cells, and active leaflet calcifications are likely to be involved in valve compartment.

Along with valve alterations, the AS-associated chronic hemodynamic stress causes left ventricular (LV) remodeling due to cardiomyocyte hypertrophy and increased production of extracellular matrix (ECM) proteins such as collagen, elastin, and glycosaminoglycans in the myocardium.

Actually, failure of the aortic valve to completely open establishes an abnormally high-pressure load upon the left ventricle. Despite the etiology causing AS, the consequent pressure overload results in the manifestation of two distinct but overlapping processes.[45] The first process is characterized by concentric LV hypertrophy (LVH).

The increased wall thickness and mass, as demonstrated by the Law of Laplace, tries to limit the increase in wall stress created by the AS-induced pressure overload state.[23456] The second process involves the myocardial ECM and leads to progressive myocardial fibrosis (MF), impairments in diastolic filling and reduced ventricular compliance.[78] It is this second phase of progressive LV MF that determines the progression of LV diastolic dysfunction and eventually leads to the presentation of the signs and symptoms of heart failure.[8910]

Simultaneously, MF contributes to systo-diastolic alterations and arrhythmogenicity, affecting patients’ prognosis and quality of life after valve replacement. Impairment in longitudinal function is the first sign of systolic impairment in AS, preceding ejection fraction (EF) reduction.

Anyway, to complicate the “puzzle,” we should not forget that in AS, the paradigm “Pressure overload - LV remodeling - myocardial hypertrophy - interstitial and later replacement fibrosis” remains still not so definite: indeed, there is a wide variation, independent from the stage of the disease (especially if we consider only valve area). Thus, some patients with severe AS have normal ventricular structure and no/mild fibrosis (10%–30%) whereas patients with only moderate AS may have extensive hypertrophy and large amount of fibrosis.

A NEW PARADIGM IN AORTIC STENOSIS: INSIGHTS FROM NEW BIOMARKERS - ROLE OF microRNAs

Aortic valve replacement (AVR), either surgical or via transcatheter aortic valve implantation (TAVI), is the only treatment that improves survival.[11] AVR is indicated in patients with severe AS who develop symptoms and/or LV systolic dysfunction (LV EF <50%). Nowadays, with the net improvement in surgical outcomes, the major management challenge is planning the correct timing of AVR. So far, biomarkers could be very useful for risk stratification in asymptomatic patients and for identification of determinants of prognosis in symptomatic patients. The key point, due to the well-known long-standing pathophysiology of the disease, is to identify patients with asymptomatic severe AS who may benefit from early AVR. As mentioned before, at earlier stages of the disease, LV EF often remains normal even though the alterations in LV myocardial structure and function can already be irreversible.

In aortic valve disease, early detection and modification of the mechanisms of disease progression could limit valve degeneration and related clinical complications.[12]

Biomarkers that reflect the degree of severity of valve disease and early LV dysfunction would be highly valuable. Lately, a big focus has been made on brain natriuretic peptides.[13] However, additional biomarkers are still required to identify early stages of disease progression in asymptomatic patients, and/or to better define the correct timing of surgery.

MicroRNAs (miRNAs) might potentially fulfill this role.

miRNAs are small noncoding RNA of about 21-nucleotide-long that regulate gene expression at the posttranscriptional level.[14] More than 1000 miRNAs have been identified in humans, possibly targeting ~ 30% of human genes.[15] Target mRNAs and biological roles have been assigned to only a few miRNAs. miRNA signatures have been associated with a wide panel of human diseases.[16] In the cardiovascular field, circulating miRNAs are currently investigated as biomarkers in acute myocardial infarction and heart failure.[17]

Cellular source of miRNAs can be diverse, including cardiomyocytes, endothelial cells, platelets, and fibroblasts, and release can occur in a disease-specific manner.[18] If the release of miRNAs into the circulation is a passive or an active process remains unknown. Circulating miRNAs can be delivered to recipient cells and modify their protein expression pattern.[19] Circulating miRNAs are more stable than mRNAs and seem to be protected from endogenous ribonuclease activity.[20] miRNAs have been involved in fibrosis, targeting ECM structural proteins or enzymes involved in ECM remodeling, in pro-fibrotic transforming growth factor-b signaling pathways, and connective tissue growth factor. They also affect epithelial-to-mesenchymal transition, induce myofibroblast proliferation, and their resistance to apoptosis [Table 1].

Table 1

miRNAs profiles: Left ventricular compartment studies

Origin	Dysregulated miRNAs†
LV* intraoperative biopsies[21]	Upregulated: 133a
LV* intraoperative biopsies[22]	Upregulated: 21
LV* intraoperative biopsies[23]	Downregulated: 1

*LV=Left ventricle, †miRNAs=MicroRNAs

Different studies compared the levels of candidate circulating miRNAs in AS cases versus controls, with the aim to identify AS-specific miRNA [Table 2]. Seven miRNAs were described to be differentially expressed in the plasma of AS patients (miR-21, miR-133, miR-1, miR-378, miR-210, miR-22, and miR-122). In these studies, high expression of miR-21 correlated with mean transvalvular gradient and LV fibrosis. miR-1 was associated with LVH and correlated with levels of soluble heart-type fatty acid-binding protein-3, a lipid-binding protein, and main target of peroxisome proliferator-activated receptor. miR-133a predicted LVH reversibility 1 year after surgery. Preoperative levels of circulating miR-133a were significantly higher in the cohort of AS patients who normalized LV mass after pressure overload release compared with those who maintained residual hypertrophy [Figure 1].

Table 2

Circulating miRNAs in aortic stenosis: Clinical studies

Patients	Findings
75 AS* patients requiring AVR† versus 32 surgical controls[22]	High expression of miR-21†† correlates with mean transvalvular gradient and LV*** fibrosis
5 AS* patients before TAVI‡‡ versus healthy controls[23]	Decreased miR-1†† correlates with increased soluble FABP3 in AS* patients upon LVH**
112 patients with moderate to severe AS* versus 40 healthy controls[24]	Lower levels of miR-1††, miR-133a††, and miR-378†† in AS* patients miR-378 levels correlate with LV*** mass index Independent predictor of LVH** in AS*
57 patients with moderate to severe AS* versus 10 healthy controls[25]	Increased miR-210†† levels in AS* patients comparable to increment in NT-proBNP levels miR-210 levels associate with higher
74 AS* patients requiring AVR†[26]	miR-133a†† as a positive predictor of the hypertrophy reversibility after surgery
94 severe AS* patients (with or without CAD‡) versus 101 controls[27]	miR-22-3p†† is upregulated in AS*, while miR-382-5p†† was downregulated, only in patients with CAD‡ miR-21-5p†† levels is higher only in AS without CAD‡ Only miR-21-5p†† and miR-382-5p†† levels correlate weakly with measures of disease severity
23 patients with HNCM§, 28 HOCM|, 47 AS*, 22 healthy controls[28]	miR-29c is specifically upregulated in aortic stenosis
38 patients with severe symptomatic AS and preserved EF[29]	miR-21†† was expressed both in myocytes and interstitial tissue, but it resulted significantly more expressed in fibrous tissue. miR-21†† levels may reflect the presence of significant myocardial fibrosis

*AS=Aortic stenosis, †AVR=Aortic valve replacement, ‡CAD=Coronary artery disease, §HNCM=Hypertrophic non-obstructive cardiomyopathy, |HOCM=Hypertrophic obstructive cardiomyopathy, ***LV=Left ventricle, **LVH=Left ventricular hypertrophy, ††miR=microRNA, ‡‡TAVI=Transcatheter aortic valve implantation. EF=Ejection fraction

Figure 1

Overview of microRNAs function in the development of left ventricle remodeling and myocardial fibrosis. Dysregulated microRNAs expression has been described in fibroblasts and myocardium. Changes in myocardial and plasma levels of miR-1, miR-21, and miR-133b and 29a are correlated

Another study compared patients with hypertrophic nonobstructive or obstructive cardiomyopathy with AS patients and described specific upregulation of miR-29c in AS.[28]

In these studies, myocardial and plasma levels of miR-1, miR-21, and miR-133a correlated directly, supporting the myocardium as a relevant source of these miRNA. Villar et al. found elevated levels of miR-21 in AS patients compared to controls. Patients with LVH due to AS showed reduced levels of miR-1, which was restored after TAVI.[222326]

So far, current knowledge allows us to assert that miRNAs might represent important regulators of valvular disease development [Table 3] and available data suggest that distinct miRNAs are dysregulated in aortic valve diseases, supporting different underlying pathophysiological mechanisms.

Table 3

miRNAs implicated in left ventricular remodeling

miR	Model	Findings
miR-1||	LV§ biopsies from AS* patients	Myocardial miR-1 expression was decreased whereas the circulating sFABP3 level was increased in AS* patients compared with healthy patients miR-1 modulates the cellular and secreted levels of FABP3, which may represent an indirect biomarker in the plasma, reflecting the cellular activity of miR-1[23]
miR-21||	LV§ biopsies from AS* patients	LV§ myocardium from AS* patients exhibits overexpression of DICER1‡ mRNA that is directly related to the expression of TGF-1***, and its effectors SMAD2, and SMAD3, and to pre-miR-21[30]
miR-21||	LV§biopsies from AS* patients	miR-21, its targets, and effectors of TGF-b*** signaling predict the variance of myocardial collagen miR-21 plays a role in myocardial fibrosis[22]
miR-133a||	LV§ biopsies from AS* patients	Combined myocardial miR-133a|| and clinical parameters predict LV§ mass normalization 1 year after AVR†[21222326313233]

*AS=Aortic stenosis, †AVR=Aortic valve replacement, ‡DICER=Double-stranded RNA-specific endoribonuclease, §LV=Left ventricle, ||miR=MicroRNA, ***TGF=Transforming growth factor

With this in mind, next studies should clarify this issues, in particular, including larger patient cohorts with controls. This to better identify the diagnostic (stage related), therapeutic, and most importantly, prognostic implications. miRNA could be included in a multimarker approach, combined with advanced imaging modalities (speckle tracking imaging as the first option) to assess small changes in myocardial texture. Hypothetically, this could contribute to tailor diagnostic approach to the single patient, refining older strategies, and conventional approaches.

OUR EXPERIENCE: MicroRNAs AND GLOBAL LONGITUDINAL STRAIN IN THE EVALUATION OF THE AORTIC VALVE DISEASE

The evaluation of patients with AS is generally limited to the assessment of flow-dependent parameters (velocity; gradients) that reflect only the “valvular side” of the pathology, disregarding LV components of disease. LV deformation by two-dimensional speckle tracking echocardiography (2D-STE) has been shown to allow a better assessment of cardiac contractile function than traditional parameters (i.e., EF), giving the chance to assess the presence of subtle alterations of LV systolic performance. In particular, regional and global longitudinal strain (GLS), showed a better and earlier diagnostic power over EF, becoming reference indicators for the precocious assessment of subclinical LV functional impairment.

In our opinion, the evaluation of both miRNA (as a biomarker) and GLS (as a functional marker) might allow an integrated assessment of the pathophysiological relationship between MF and adverse LV remodeling. With these considerations in mind, in our pilot study, we aimed at assessing:

The presence of MF (endomyocardial biopsy) both with gold standard histologic method and with an advanced laser microdissection (LMD) methodology (tissue miRNA-21), in patients with severe AVS

The association between plasmatic and tissue pool of miRNA-21

The relationship between 2D-STE parameters, MF, and plasmatic/tissue miRNA-21 expression levels.

This to develop a noninvasive algorithm for detection of myocardial fibrotic burden and systolic functional impairment.

We included 38 consecutive patients with severe symptomatic AS and preserved EF candidate to surgical AVR. Patients underwent laboratory analysis and transthoracic echocardiography (M-Mode, 2D, Doppler, tissue Doppler imaging [TDI], 2D-STE). Twenty-eight patients were finally submitted to AVR, 23 of whom underwent intraoperatory basal interventricular septum biopsy to evaluate MF and tissutal levels of expression of miRNA-21. Conventional echocardiography by transthoracic examination was performed.[3132]

Assessment of LV GLS was performed using 2D-STE. For dedicated septal analysis, a focused region of interest was traced specifically for interventricular septum. We assessed septal longitudinal systolic strain (SSL), systolic (SL-Sr) strain rates, and early diastolic (SL-SrE) strain rates. We then averaged measure from anterior and inferior septum. In 23 patients undergoing surgical AVR, concomitant intraoperatory basal left side interventricular septum biopsy was performed to assess MF.[33] Antigen retrieval was performed microwaving sections.

Peripheral blood samples for microRNA were collected using specific tubes. For twenty patients’ biopsy specimens, LSMD was performed. The RNA from both plasmatic and tissues samples was purified. MicroRNAs were reverse transcribed and then cDNA obtained was preamplified before real-time polymerase chain reaction analysis of miRNA, so miRNA 21 expression was calculated.

Even if conventional indices of global systolic function were preserved (EF, fractional shortening), more sensitive parameters of longitudinal function (e.g., MAPSE; TDI) were reduced when compared to normal ranges.

Speckle tracking analysis revealed a significant impairment of global longitudinal deformation parameters (GLS). Septal subanalysis showed a higher impairment of deformation indices and a significant direct relationship of SSL with stroke volume.

A variable amount of MF (with the absence of inflammatory cells) was a common finding in patients who performed biopsy.

MF showed a significant inverse relationship with deformation indices (GLS: R2 = 0.30 and P = 0.02; SSL: R2 = 0.36 and P = 0.01; SL-Sr: R2 = 0.39 and P < 0.001; SL-SrE: R2 = 0.35 and P = 0.001). While miRNA-21 was expressed both in myocytes and interstitial tissue, it resulted significantly more expressed in fibrous tissue (P < 0.0001), and it was inversely related to septal and global longitudinal deformation (SSL: R2 = 0.32 and P = 0.01; GLS: R2 = 0.34 and P = 0.008). Plasmatic miRNA-21 concentrations demonstrated a significant direct relationship with whole MF (R2 = 0.31; P = 0.001) and interstitial miRNA-21 compartment (R2 = 0.36; P = 0.001).

A significant and strong positive relationship between MF and plasmatic miRNA-21 was found, also after weighting for cardiac remodeling (assessed as left ventricular mass index: R2 = 0.50; P = 0.0005) and LV function parameters (SSL R2 = 0.35; P = 0.006).

At ROC analysis, a plasmatic miRNA-21 value >2.4552 showed the best accuracy (Sensitivity 64.29%; specificity 100%; AUC 0.81; P = 0.001) for discriminating patients with significant MF (described as equal or more than 10% of the specimen).

Thus, to summarize, the principal findings of our research were:

A direct link between regional and global LV myocardial strain and invasively measured MF (gold standard)

A pathophysiological role in AVS for miRNA-21. In particular, the interdependence between textural miRNA-21 and fibrogenic stimulus induced by an abnormal LV pressure overload

Circulating miRNA-21 (biomarker) levels may reflect the presence of significant MF.

CONCLUSION

The present short review shows that an integrated approach using echocardiographic and plasmatic miRNAs analysis in the complete evaluation of severe AS patient could have a role in the next future, essentially for a better characterization of the remodeling process at macroscopic (ultrasonic speckle tracking analysis) and biohumoral level (mRNAs), combined with the standard and well-defined evaluation of the complex aortic valve apparatus status. Furthermore, this integrated approach could help in better stratifying those patients that currently fall in a diagnostic “gray zone” of AVS severity.[29]

Large prospective studies involving patients with AVS, also exploring the power of other biomarkers such as other types of mRNAs or other promising biomarkers (i.e., ST2) are needed to better analyze the effective diagnostic and prognostic value of this cardiac imaging and biohumoral integrated approach.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.