Can miRNAs be useful biomarkers in improving prognostic stratification in endometrial cancer patients? An update review

Abstract Endometrial cancer (EC) is the most common gynecological cancer, with annual incidence rates in Western countries ranging between 15 and 25 per 100 000 women. About 15% to 20% of patients with EC have high‐risk disease and follow an aggressive clinical course. Unfortunately, the assessment of histologic parameters is poorly reproducible and conventional clinicopathological and molecular features do not reliably predict either the patient's response to the available treatments or the definition of personalized therapeutic approaches. In this context, the identification of novel diagnostic and prognostic biomarkers, which can be integrated in the current classification schemes, represents an unmet clinical need and an important challenge. miRNAs are key players in cancer by regulating the expression of specific target genes. Their role in EC, in association with clinical and prognostic tumor biomarkers, has been investigated but, so far, with little consensus among the studies. The present review aims to describe the recent advances in miRNAs research in EC taking into consideration the current classification schemes and to highlight the most promising miRNAs. Finally, a perspective point of view sheds light on the challenges ahead in the landscape of EC.

ovarian syndrome and all causes of hyper-estrogenism not balanced by progesterone/progestin presence. This frame describes a multifactorial and heterogeneous disease, in which the prognosis is based on surgical and pathological factors: International Federation of Gynecology and Obstetrics (FIGO) stage, grading and histotype, depth of myometrial invasion, lymphovascular space invasion (LVSI) and lymph node metastases (LNM).
Currently, the stratification risk system appears to be insufficient and inadequately informative in current clinical practice and often it presents challenges in identifying the most appropriate therapeutic approach. Eventually, histopathological parameters used to identify risk factors are not always easily reproducible, particularly in high-grade carcinomas with intratumoral heterogeneity. All these factors suggest that EC is inappropriately treated due to subjective interpretation of clinicopathological characteristics even by experienced clinicians. [4][5][6][7][8][9][10][11][12] A new classification based on the molecular characteristics of the EC was proposed in 2013 by The Cancer Genome Atlas (TCGA), however this classification is complex to reproduce in a clinical laboratory and results in a poor clinical translation, despite the surrogate molecular analyses proposed independently by the Vancouver and European Groups. 13 In this context, the identification of novel biomarkers remains an unmet clinical need and an appealing opportunity is represented by the small non coding RNAs (ncRNAs), in particular, micro-RNAs (miRNAs). In the last decade, the research on miRNAs in EC have increasingly grown; nevertheless, the integration of miRNAs results with the proposed clinical and molecular classifications is very limited, leaving an important knowledge-gap.

| Histological classification
For years, this disease has been simply divided and trivialized into two main groups: estrogen-dependent (Type I) and estrogen-independent (Type II). Type I, low and intermediate grade endometriod ECs (EECs), is the most common (70%), it is associated with hormone receptor positivity 4,5 and, generally, it has a favorable prognosis. Type II includes high grade EECs and nonendometrioid subtypes such as serous (SEC), clear cell (CCC), carcinosarcomas (CS) and undifferentiated EC. These tumors are estrogen-independent, not associated with obesity and with a poor prognosis. 4 However, it is now clear that this simple dichotomous division is not adequate to represent the complex heterogeneity of this tumor. Generally, EC is considered a cancer with a favorable prognosis, therefore, far from the well-defined diagnostic-therapeutic structure of ovarian carcinoma, rarer and more lethal. But over time, it has been realized that not all ECs have a favorable prognosis, and some of them resemble ovarian carcinoma. 6 This happens not only for Type II but also for some Type I tumors that sometimes show a surprisingly unexpected aggressive behavior.

| Classification based on the ESMO risk
The ESMO has identified specific prognostic factors (i, FIGO stage; ii, grade; iii, depth of myometrial invasion; iv, LVSI) based on which clinicians stratify patients into four distinct risk groups (low, intermediate, with an intermediate prognosis. 13 14,15 Similarly, the TransPORTEC has recognized four prognostic subsets, diagnosing first POLE mutant tumors, then MMRd tumors, p53-mutant tumors and ECs with NSMP. 16 In both the cases, after ECs stratification based on these molecular features, prognostic signatures emerged. This prompted to a diagnostic algorithm that includes molecular criteria, resulting in a novel, more unbiassed and clinically meaningful EC classification. The clinical and prognostic relevance of these molecular subgroups has been corroborated in several studies 14,[17][18][19] and the most recent ESGO guidelines recommend an integration of standard clinicopathological parameters and new risk profiles, suggesting molecular characterization in all endometrial carcinomas, especially those at high risk. 20 However, some histopathological parameters, independently by the molecular subtype, as stage and LVSI, do not have a molecular surrogate and are still critical in the pathological evaluation. Ideally, molecular and clinicopathologic prognostic classification schemes work better together. Unfortunately, the proposed molecular system remains difficult to apply to the routine clinical management as it involves laboratory equipments to carry out definite molecular and pathological analyses which, often, are not available in the peripherical facilities. Moreover, recent studies proposed additional classes to be implemented in the recent classification in order to further stratify the risk of relapse.

| MiRNAs
Since the ENCODE project highlighted that about 75% of the genome is transcribed in RNAs and only 3% is represented by gene encoding for proteins, the research has made several advances and the ncRNAs have been recognized as key players in many biological processes. 21,22 MiRNAs, in particular, are small ncRNAs of about 19 to 24 nucleotides (nts) which play an important role in modulating the expression of their targets at the post-transcriptional level. This regulation is mediated by base pairing to a 6 -8 nts sequence of the mRNA target, with perfect or imperfect complementarity and leads to expression inhibition by mRNA degradation or translational repression; a single miRNA can modulate the expression of hundreds of mRNA targets and, conversely, a mRNA target may be regulated by multiple miRNAs. [23][24][25][26] In the last two decades, the research has progressively characterized miRNAs function and mechanism of action to finely regulate the target genes. Today, it is well recognized that miRNAs play a pivotal role in several processes, such as cell growth development, cell cycle, apoptosis and many others. Compelling evidences have established that miRNA expression is dysregulated in human diseases, including cancer. Of note, miRNAs may act as oncogenes or tumor suppressors and, depending on cellular context, the same miRNA may behave in different ways. Therefore, deregulation of miRNAs can affect the hallmarks of cancer, including sustaining proliferation, evading apoptosis and resisting cell death, promoting invasion and metastasis and inducing angiogenesis. 27,28 Over the past decade, with the research advances, a plethora of deregulated miRNAs have been progressively described in a wide range of solid and liquid malignancies; this highlights the importance of these epigenetic regulators in fine tuning the biological processes. 24,25

| MiRNAs IN EC
In recent years, many studies have identified different genes and miRNAs as potential biomarkers in EC. [29][30][31][32][33][34][35] However, in most of these, especially when focused on miRNAs, deregulation was observed by comparing tumor with normal counterpart or healthy tissue. 36 In this context, among others, a complete overview is presented by Srivastava et al, who reviewed the miRNAs differentially expressed in EC with respect to normal endometrial tissue. 37 On the contrary, the studies that analyzed miRNA expression taking into account other parameters, such as grade or the existing classifications schemes, particularly the TCGA one, are limited and with small consensus.
Given these premises, the aim of this narrative review is to provide an overview of the miRNAs that have been identified as significantly associated with EC molecular and clinical features and that could be potentially integrated in the available classifications.
To this purpose we searched for papers analyzing miRNA expression in EC in relation to prognostic and molecular classifications; reports comparing miRNAs between EC and normal tissue were not considered because out of our scope.
The papers included in this revision are summarized in Table 1, whereas Table 2 reports the main role and potential targets of miRNAs described in the above-mentioned papers. In our review, first, we focused on studies that analyzed tissue miRNAs and prognostic factors and, subsequently, on the ones investigating the potential association between circulating miRNAs and prognostic factors, based on the current classifications.

| MiRNAs associated with stage and grade in EC
According to the FIGO 38  An interesting miRNA family that seems to be correlated with EC metastatic potential is the miRNA-200 family. 39 This family, one of the best characterized, consists of five miRNAs (miR-200a, miR-200b, miR-200c, miR-141 and miR-429) which may negatively regulate expression of ZEB-1/2, that play a critical role in epithelial to mesenchymal transition (EMT). [40][41][42] In EC, this biological process is particu- . The results showed that miR-205 was significantly higher expressed in early stage EC patients with respect to advanced stage tumors. In addition, poorly differentiated (Grade 3) tumors presented lower expression compared to moderately differentiated. 48 Interestingly, the independent study by Chung et al, previously mentioned, showed that aberrant expression of miR-205, beside miR-200a, was significantly correlated with advanced stages. 46 On the contrary, contrasting findings were reported by Tsukamoto and collaborators, who showed that miR-205 level was higher in Grade 2 and 3 tumors compared to Grade 1 ECs. 49 Another miRNA potentially involved in EMT and recently explored in EC is miR-210. 50 This has shown abnormal expression in several cancers, 51  T A B L E 2 Potential role or targets of miRNAs proposed by the analyzed papers miRNA ID Reference describing the miRNA Potential role or targets of miRNAs let- 7c-5p  Kalinkova et al, 2020 58  NRAS, PIK3R5, TP53, AKT2, CCND1, APC2, PIK3CA   let-7g-5p  Kalinkova et al, 2020 58  BRAF, NRAS, KRAS, MLH1, TP53, CCND1, CTNNA1, MYC, MAPK1,  PIK3R5, AKT2, APC2, PIK3CA   miR-125b-5p  Kalinkova et al, 2020 58  ERBB2, RAF1, AXIN1, TP53, CTNNB1, CTNNA1, AKT1, PIK3CB, 54 The selection of this miRNA was due to its potential suppressive effects on cancer initiation and progression. 55 the TCGA molecular classes. In particular, miR-499a-5p resulted upregulated in the NSMP EC patients harboring CTNNB1 mutations; the correlation was also corroborated in an independent, larger EC cohort extrapolated by the TCGA cohort. Furthermore, by combining the miRNA expression with the CTNNB1 mutational status, the authors identified a subgroup of NSMP patients with better OS, and miR-499a-5p resulted an independent risk factor of death. 65

| MiRNAs associated with tumor recurrence
As previously mentioned, Type I EEC is the most common histologic subtype. Usually the majority of endometrioid EC patients are considered to be at low risk, when diagnosed at an early stage, with no benefits from further treatment after surgery. [66][67][68] Though, a subset of these patients shows a higher risk of recurrence and poor overall prognosis. Unfortunately, to date, no accurate tools to identify these high-risk patients are available and their management remains challenging. Given that, identification of novel biomarkers able to clinically drive the correct patients' selection remains an unmet clinical need.
For this reason, researchers have looked with interest to the miRNAs.
In this regard, Devor and coworkers analyzed 54 recurrent and nonrecurrent ECs cases from the Gynecologic Oncology Group (GOG) Study-210. 69 The study cohort was well balanced, composed of

| Circulating miRNAs in EC
In the last years, we have witnessed significant advances in the socalled liquid biopsy from a diagnostic and prognostic point of view.
On the same wave, circulating miRNAs have been deeply characterized in many cancer types and EC has not been left out. However, with the aim to outline noninvasive approaches for diagnosis, the majority of the studies focused on EC have, so far, assessed the miRNA levels in plasma or serum samples between oncological patients and healthy individuals [75][76][77][78] (see Table S1 for a summary Based on the data reported in our work, the most appealing miRNAs in EC belong to the miR-200 and the miR-34 families. So far, indeed, multiple research groups have showed the association of these miRNAs with EC characteristics. These correlations could be due to their involvement in the EMT process which is known to play an important role in EC progression, metastases and recurrence and being one of the cancer escape routes to medical treatments; Figure 2 depicts the main miRNAs involved in EMT in EC. However, additional studies are strongly warranted in order to elucidate the role and to clarify if one or more members of miR-34 and miR-200 families may represent reliable diagnostic, prognostic biomarkers or potential therapeutic targets.
In conclusion, besides the wide efforts done in characterizing miRNAs in EC, their utility is currently still scarce, and a global effort should be considered as mandatory to achieve a clinical translation of tissue miRNAs and, possibly, "liquid miRNAs" into the clinic.