Recently, the current knowledge regarding the general miRNA expression changes in PAH has been thoroughly summarized (3). In this article, we have focused on three clinical problems in patients with PAH, which could be solved using miRNA-based biomarkers.

Firstly, differentiation of the etiology of PAH is difficult. In registries, about 50% of patients with PAH have idiopathic, heritable or drug-induced PAH. It is important to increase our knowledge in order to minimize the diagnosis of idiopathic PAH. The remaining cases of PAH, described as associated PAH, can be caused by various conditions, such as connective tissue disease (CTD), human immunodeficiency virus (HIV) infection, schistosomiasis or portal hypertension (4). The leading cause in the subgroup of associated PAH is CTD, mainly systemic sclerosis (5). In systemic sclerosis, it is difficult to determine whether increased pressure in pulmonary artery is caused by pulmonary fibrosis (ESC group 3), small vessel disease (ESC group 1 = PAH) or left heart failure associated with left ventricle diastolic dysfunction (ESC group 2). Differentiation of the cause is extremely important due to the different treatment. Congenital heart disease (CHD) can also play a role in the development of PAH. It is estimated, that 10% of adults with CHD may also have PAH (6). Currently, there is no specific blood-based biomarker for PAH, although numerous markers have been tested (7), such as inflammatory markers [growth differentiation factor-15 (8), osteopontin (9), neutrophil to lymphocyte ratio (10), stem cell growth factor β (11)], heart function markers (B-type natriuretic peptide, amino-terminal pro-B type natriuretic peptide or red cell distribution width) or oxidative stress related biomarkers [F2-isoprostanes (12) or oxidized lipids] (9, 13).

Second, it is challenging to determine the severity of PAH, and thus to qualify patients for an appropriate treatment. Although some blood test might help to identify the PAH etiology, they are not useful in diagnosing or assessing the clinical advancement of the disease (7). Hence, the diagnostic workup currently includes repeated RHC, which is an invasive procedure with a risk of complications estimated according at about 1%. Altogether, PAH requires novel tools to evaluate the clinical severity of the disease.

Third, the results of PAH treatment remain unsatisfactory (7). As stated in the guidelines of the European Society of Cardiology, there are three pathways underlying the pathogenesis of PAH: the endothelin, nitric oxide (NO) and prostacyclin (PGI₂) pathways (7). Even though treatments targeting each one of them, including endothelin receptor antagonists (ERA), phosphodiesterase type 5 inhibitors (PDE-5) and PGI₂ analogs are well established in clinical practice, the functional limitation and survival of patients leaves room for substantial improvement (7). Hence, new therapeutic targets in PAH are urgently needed.

MicroRNAs(miRNAs) are small non-coding RNAs that regulate post transcriptional gene-expression. MiRNAs can induce mRNA degradation and translational repression by interacting with the 3′ untranslated region (3′ UTR) of target mRNAs, but their interactions with other regions of mRNAs have also been reported. In some circumstances, they can also activate translation or regulate transcription. However, miRNA interaction with their target genes is very dynamic and dependent on many factors, such as subcellular location of miRNAs, the abundancy of miRNAs and target mRNAs, and the affinity of miRNA-mRNA interactions (14). MiRNAs are found in all eukaryotic cells across the species. Over the last few years, their role in the progression and development of numerous diseases has been extensively studied, since they hold enormous potential for prognostic, and diagnostic biomarkers as well as therapeutics (15–19). Studies showed that miRNAs are also promising novel biomarkers in PAH due to their activity in various molecular pathways and processes underlying PAH (3, 20). The lack of biomarkers in PAH enabling the differentiation of individual PAH etiologies and the assessment of the severity of PAH as well as the shortage of therapeutic targets in PAH opens a new field for the possibility of using miRNAs in these applications (3).

In this review, we present current challenges in miRNA analysis and interpretation in PAH and summarize the steps which need to be taken before clinical applicability of miRNAs in PAH.

Etiology of PAH can be characterized by unique biochemical profile of particular miRNAs expressions. Usually, changes in the concentrations of many miRNAs are observed compared with healthy volunteers. In some etiologies of PAH, the bone morphogenetic protein receptor type II (BMPR2) gene has been found mutated (42). Product of this gene transcription, BMPR2 protein is a serine/threonine receptor kinase, which is the most important mediator of vascular remodeling in PAH. The dysregulation of this receptor kinase has been found in hereditary PAH (43, 44), and downregulated in some other etiologies of PAH (42). The downregulation of BMPR2 may be caused by the action of miRNAs, which appeared to be essential regulators in the pathogenesis of PAH, hence their potential use as biomarkers in this disease as a whole, and also in particular etiologies of PAH. MiRNA was experimentally proven to regulate the expression of BMPR2 (42). An altered expression miR-125a in lung tissue samples of experimental PAH was observed. There is a possibility of evaluating expression levels of miR-125 in order to assess hereditary PAH, since miR-125 takes part in downregulation of BMPR2, which has been seen in this particular PAH etiology. Potential role of miRNA as a regulator of BMPR2 is presented in Figure 1. Importance of BMPR2 signaling pathway in PAH pathogenesis was also proven in a recent study with sotatercept. It is a first-in-class therapeutic fusion protein that targets an imbalance in activin–growth differentiation factor and BMPR2 pathway signaling (45). Twenty four-weeks treatment with sotatercept reduced pulmonary vascular resistance among patients with pulmonary arterial hypertension who were receiving stable background therapy, including prostacyclin infusion therapy. Concordant improvements from baseline in 6-min walk distance (6MWD) and NT-proBNP levels were also observed. 6MWD is a parameter used to assess functional capacity. It is widely used in clinical trials to compare the effects of treatments (46).

MiRNAs could be used to evaluate the hemodynamic and clinical advancement of PAH. There are studies showing that the concentration of miRNAs in lungs of patients with PAH is different compared with samples from healthy controls (20). Correlation between hemodynamic severity of PAH and plasma concentration of specific miRNAs could be crucial in a diagnostic process and could also provide essential prognostic information (20).

In the perfect situation, level of mRNA is constant and the level of the protein produced by it is lowered or increased due to a disturbed level of miRNA (61). However, various mechanisms influence mRNA expression. The long non-coding RNAs (lncRNA), which plays a role not only in regulating post-transcriptional mechanisms, such as miRNA, but also transcriptional mechanisms, may be also responsible for changes in mRNA expression (3, 16, 62, 63). The role of lncRNA has been demonstrated, inter alia, in the pathogenesis of PAH resulting from PASMC and PAEC dysfunction (3, 61). Hence, in cases of PAH where the mRNA level is normal and the level of the protein produced from this mRNA is abnormal, the miRNA biomarkers may be partially insensitive due to overlapping factors, such as lncRNA (64). LncRNAs are >200 base pairs in length and are abundantly present in the genome. Almost 27,000 lncRNAs have been identified in the human genome already (65). They are differentially expressed in key pulmonary vascular cells such as PAECs and PASMCs, which regulate a range of cellular and biological processes and contribute to pulmonary vascular remodeling and PAH pathogenesis (64). Summary of lncRNA involved in PAH is presented in Table 2.

One of the challenges regarding the use of miRNAs as biomarkers of human diseases is deriving conclusions from studies conducted in animal models (60). Although these models show similar miRNA expression for healthy animals as healthy humans, miRNA production for diseased individuals differs in animals and humans (60). These differences are due to different pathophysiological mechanisms of PAH in animals, which are also phenotypically different from human PAH in many aspects (60). However, animal studies are necessary, because the small number of PAH patients makes it difficult to develop cohort studies. In many cases, there is also a need to perform invasive procedures (60). The triggers of PAH used in animal models include, but are not only limited to, administration of monocrotaline or chronic hypoxia (60). The challenge is therefore to find the PAH trigger that most closely matches the cellular and molecular changes present in human PAH.

Another challenge for the use of miRNA as a biomarker is the likely existence of multiple PAH etiologies with different miRNA expression in one patient (3, 66). These etiologies differ in the number of inflammatory and hypoxic stimuli leading to the formation of PAH and causing different expression of miRNAs (42). A suggestion for the existence of different etiologies of PAH in patients is that the animal models of PAH which are otherwise induced, show different miRNA expression (3, 27). Thus, the usefulness of many biomarkers may depend on the etiological factors of PAH.

Furthermore, miRNA expression may also depend on the degree of progression of PAH in the patient. There are evidences, that it depends on the development of right ventricular (RV) hypertrophy or RV failure (3). Some studies suggested that different miRNAs may be detected if there is the presence of RV hypertrophy or failure (28, 29). Thus, the detection of miRNAs, which are produced by dysfunctional heart muscle, should take this phenomenon into account.

There are numerous factors influencing the accuracy of miRNA analyzes. A critical evaluation of such factors is of paramount importance in order to avoid misleading results and to allow the comparability of different analyzes. For example, in RNA isolation methods, maximum efficiency, reproducibility and reliability are of great importance, because, e.g., even minimal co-extraction of inhibitory factors can significantly affect the measurement results. Various detection methods (high-throughput sequencing, real-time PCR, microarrays) are used to quantify miRNA levels. Their application should be determined for specific indications to minimize variation between studies. Standardization method should be standardized and reproducible isolation and detection protocols applied; otherwise, comparability is much more difficult (67). In addition to these analytical aspects, there are parameters influencing the levels of miRNAs in the analyzed biomaterial. Drugs such as statins, antiplatelet drugs and heparin can have a confusing effect on miRNA measurements. For example, antiplatelet drugs reduce the amount of freely circulating thrombocyte-derived miRNAs, and heparin influences the polymerase chain reaction (68). Finally, there are differences in the results of miRNA quantification between cellular and extracellular sources of biomaterial (69).

In a study of 2,391 individuals expression of miRNAs measured by high-throughput RT-qPCR isolated from both whole blood and were compared. There was only modest association between cellular and extracellular miRNA expression in this large study sample (69). These results demonstrate that miRNA expression from cellular vs. acellular sources may be divergent. MiRNA concentration from different human sources (e.g., whole blood, plasma) should not be used interchangeably as biomarkers of disease (69).

The first option is to use miRNA as a diagnostic marker (3, 20). So far, it has not been possible to find an ideal biomarker for PAH, i.e., one that is non-invasive, easy of measure, cost-effective, and has high sensitivity and specificity (70, 71). There is a great hope in finding such a marker among the miRNAs (56). For this purpose, the problem of different fragments and levels of miRNAs in experimental studies caused by the lack of their standardization should be solved (3, 56). It is caused by the lack of systematization in the procedure of taking, storing and processing samples; lack of standardization in the methods of miRNAs diagnosing and lack of indications for their use; lack of standardization in miRNA quantification protocols; rare sampling in the studies and lack of described concomitant disease symptoms (3, 56, 70, 72). After solving the above problems, it would be possible to summarize the results of many different studies and use the obtained results to create studies on a large samples (56). These studies would justify the use of specific miRNAs as biomarkers. Machine learning might be useful for selection of the most promising miRNAs. Machine learning as a field has progressively improved our ability to find relevant features in large and high-dimensional data sets collected from genomic studies (73). Supervised machine learning methods have been used successfully to develop classifiers for disease diagnosis, as well as to identify potential disease biomarkers (74). In a recently published study it was shown that using a consensus of four different supervised machine learning feature selection techniques can help us to identify miRNAs associated with PAH (72). MiR-187-5p and miR-636 were selected as a candidate biomarkers that might be associated with PAH progression, however this method requires further studies and validation (72).

As mentioned earlier, miRNA concentrations in RV hypertrophy and RV failure vary (27, 28). This suggests that during the progression of RV overload caused by, inter alia, PAH there are some changes in miRNA expression. More research is needed to see if we can predict PAH progression from changes in miRNA levels even before structural changes occur, especially in patients with high risk of developing PAH.

The final step of identification of perfect biomarker would be to perform a cost-effectiveness analysis to determine whether the ratio of new information obtained from such a research and its price is satisfactory to use such a miRNA for the diagnosis of PAH (56).

The future prospects of the miRNA are also apparent in their use in the treatment of PAH. Mainly, such therapy would be based on the use of miRNA analogs to increase the amount of miRNA that is decreased in PAH, or miRNA inhibitors to decrease the amount of miRNA whose expression is increased (20, 56). So far, the experiments in animal models show promising results (3). Before they can be used clinically, however, some problems have to be solved. These include: low drug stability, off-target effect and drug delivery to the selected tissue (3, 20, 56, 75).