Heart failure (HF), a terminal stage of most cardiovascular diseases, is a major cause of hospitalizations and mortality worldwide (1, 2). The incidence of HF is raising rapidly because of increasing risk factors including hypertension and diabetes, and it is more common in elderly people over 80 years old (3, 4). Currently, the serum brain natriuretic peptide is the only routinely used biomarker for HF recommended in guideline (5), but its clinical value is still uncertain. Therefore, the predictive biomarkers are urgently needed for the early diagnosis of HF. In recent years, a growing number of researches have reported transcriptomic changes are associated with the pathophysiological mechanism of HF (6). Circulating microRNAs (miRNAs) gained significant interest as potential novel HF biomarkers because of high stability, sequence conservation, and non-invasive detection (7). However, prognostic value of microRNAs in HF is still unclear. Thus, the systematic review of miRNAs may contribute to identifying biomarker for early clinical diagnosis of HF.

MiRNAs are a class of conserved non-coding RNAs that regulate protein synthesis on the post-transcriptional level (8). MiRNAs are involved in the pathophysiology of various diseases (9–14), and the aberrant expression of miRNAs were potential biomarker for cardiovascular diseases (15, 16). Increasing evidence has demonstrated dysregulated miRNAs could alter the cellular response of cardiomyocytes, leading to cardiac concentric hypertrophy (17, 18). For example, miR-127 aggravates myocardial failure by regulating the expression of TGF-β1/Smad3 signaling pathway (19), miR-129-5p improves cardiac function in rats with chronic HF through targeting HMGB1 (20), and upregulated miR-132 improved the cardiac dysfunction by inhibiting PTEN expression (21). In general, understanding the pathophysiologic mechanisms in HF could provide the chance to develop new therapies (22, 23).

In previous studies, numerous studies have verified that the circulating miRNAs could be serve as potential biomarkers. For example, miRNA-208 and miRNA-150 have been identified as potential biomarkers for cardiac hypertrophy and remodeling (24, 25), miRNA-21 may serve as a biomarker for coronary heart diseases in the elderly patients (26). MiRNAs change in certain diseases, especially circulating miRNAs due to their stability makes them possible biomarkers in HF (27). Previous studies have reported that the dysregulated miRNAs can serve as the potential diagnostic biomarker in HF (16, 28, 29). Additionally, miRNA-19b, miR-21, miR-423-5p, and miR-92b-5p were observed to be promising biomarker for HF (16, 29–31). Given the pathogenic role of miRNAs in HF, the present study explored an association between miRNAs expression signatures and HF.

Over the past years, the dysregulated miRNAs were investigated in multiple studies, which come from different pathophysiological process, sample size, and inconsistent inclusion criteria, and the expression profiles of miRNAs in HF remain elusive (32, 33). Therefore, we aim to gather the currently available data about dysregulated miRNAs in HF, and to explore promising biomarker for diagnosis and prognosis of HF.

Over the past years, miRNAs have be observed to play key roles in the physiopathology of HF (56, 72), and there is increasing number of articles reporting on the differentially expressed miRNAs in HF (21, 29, 35). The identification of the dysregulated miRNAs may lead to the novel discovery to monitor the progression of HF. Nevertheless, a major problem with miRNA expression profiling was the inconsistency among different studies. Therefore, we reassembled the data and summarized the abnormal expression signatures of multiple miRNAs in HF through comprehensive data analysis. Finally, this systematic review identified 57 aberrantly expressed miRNAs (47 upregulated, and 10 downregulated) in all included studies. Based on further analysis, several miRNAs (miR-21, miR-30c, miR-210-3p, let-7i-5p, miR-129, let-7e-5p, and miR-622) were identified as potential biomarkers of HF.

The dysregulation of miRNAs, their target genes, are related to signaling pathways of HF, including cardiac hypertrophy, oxidative stress, cardiac fibrosis, etc. (30, 34, 35). The upregulated let-7a regulates β₁-adrenoceptors (AR) signaling pathway by suppressing β₁-AR expression in ischemia HF (75). The target genes of the top dysregulation miRNAs identified in this systematic review are all associated with HF. MiR-21 induces fibrotic process by regulating TGFβ1-Smad3 signaling (83), and the inhibition of miR-21 by specific antagomiR leads to reduced hypertrophy and fibrosis (84). Hypoxia-inducible factor-1α (HIF-1α), the target of miR-221-3p inhibits angiogenesis in HF (62). Additionally, miR-122-5p and miR-184 were proved to promote apoptosis in post-infarction HF (61). MiR-150 regulates apoptosis signaling by targeting pro-apoptotic genes and suppressing p53 activity as a major inducer of apoptosis (85).

To date, the early diagnosis of HF is still challenging, a more novel and advanced biomarker are urgently needed. According to the previous studies, miRNAs can be used as a new method for monitoring of HF. The dysregulated miRNAs (miR-19b, miR-129-5p, miR-221-3p, miR-208a, etc.) may provide non-invasive biomarkers for the diagnosis and prognosis of HF (29, 36, 62, 64). Therefore, the practicable detection methods are essential to ensure the accuracy and precision of miRNAs (86). Currently, there are the increasing concern about normalization strategies, that may contribute to the inconsistent expression data of microRNAs in PCR detection (87). The appropriate internal controls in PCR technology were required for the quantification of miRNAs, and relevant studies have verified that the spiked exogenous C. elegans miRNAs (e.g., miR-17, miR-454, cel-miR-39, and RNU6) as assay-normalization controls, did not significantly alter assay imprecision (16, 87–90). In PCR-based microRNAs detecting, the findings were robust irrespective of the way of standardization (91). Therefore, the findings in this meta-analysis were relatively reliable. However, in the future study, the rigorous experimental verification for confirming their validity with HF is necessary.

It was investigated whether miRNAs could serve as a promising biomarker of HF. Circulating miRNAs are promising diagnostic biomarkers for HF because of their high sensitivity and specificity. The increasing numbers of studies have focused on the clinical role of miRNAs. However, due to different specimens, sample sizes and regions, the current miRNAs expression characteristics were contradictive. The inconsistent results about dysregulated miRNAs may be caused by a variety of factors, including ethnic differences, the source of samples, and sample sizes. For example, plasma let-7e was upregulated (60, 66), while was downregulated in myocardial tissue (75). MiR-125a-5p was observed to be significantly downregulated (73), but upregulated in animal sample (40). Meanwhile, effects of concomitant treatment on miRNA expression in cardiovascular diseases should not be ignored. Actually, concomitant pharmacological treatment may have an important influence on circulating miRNA signatures. As reported in previous studies, significant changes in circulating miRNAs were observed in response to antiplatelet therapy (92, 93), and antidiabetic treatment increased the levels of let-7a and let-7f (94). The pharmacological treatment should be addressed and provided in the miRNA studies. Regrettably, among our included studies, few studies reported concomitant drugs, and little is known about the influence of drug treatment on the miRNA modulation. Therefore, in future studies, concomitant treatment should be taken into account when using miRNAs as diagnostic biomarker for cardiovascular diseases.

In the present systematic review, subgroup analysis was performed to explore probable sources of heterogeneity. The findings suggest that blood sample may be more informative about HF indicator. 45 dysregulated miRNAs were identified, the most down-regulated miRNAs was miR-30c, and the most upregulated was miR-210-3p. Because the circulating miRNAs are relatively stable and easily detectable by non-invasive methodology (95), circulating miR-30c and miR-210-3p, may be ideal candidate biomarkers for HF. Besides, in Asian countries, let-7i-5p and miR-195 were the most upregulated miRNAs, and miR-129 was the most downregulated one. On the contrary, in non-Asian countries, the significantly upregulated and downregulated miRNAs were let-7e-5p and miR-30c, respectively. Further analysis reveals that more miRNAs are dysregulated in humans than in animals, including miR-21, miR-423-5p, and miR-328-3p. These miRNAs may be the most suitable non-invasive biomarkers in the human’s blood for diagnosing and monitoring of HF.

To the best of our knowledge, the present meta-analysis was the comprehensive and credible to evaluate the miRNA expression signatures in HF. We combined the data of 55 articles to enhance the statistical power and the reliability of the results. However, there are still some limitations. Firstly, individual studies included relatively few patients, which limiting the strength of the conclusions. We thus conducted the sensitivity analysis to guarantee the robustness of the results. Secondly, although a comprehensive literature search was applied, some valuable research may be missing. Additionally, the biological characteristics and mechanisms of the different miRNAs in HF may differ, which may limit the applicability of the pooled results. We thus carried out subgroup analysis to increase the validity and reliability of miRNA analysis for HF.

This systematic review on miRNA expression profiling studies confirmed several valuable miRNAs, including miR-21, miR-30c, miR-210-3p, let-7i-5p, miR-129, let-7e-5p, and miR-622. These miRNAs may be used as potential non-invasive biomarkers and drug targets for HF. In the future, further investigations are required to verify the value of miRNAs in diagnostic and therapeutic approaches in HF.

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

Y-PF was the guarantors of the entire manuscript. Y-PF and N-NS contributed to the study conception and design, critical revision of the manuscript for important intellectual content, and final approval of the version to be published. J-LW contributed to the data acquisition, analysis, and interpretation. All authors contributed to the article and approved the submitted version.

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.