The Role of Cytokines in Epithelial–Mesenchymal Transition in Gynaecological Cancers: A Systematic Review

Chronic inflammation has been closely linked to the development and progression of various cancers. The epithelial–mesenchymal transition (EMT) is a process involving the acquisition of mesenchymal features by carcinoma cells and is an important link between inflammation and cancer development. Inflammatory mediators in the tumour micro-environment, such as cytokines and chemokines, can promote EMT changes in cancer cells. The aim of this systematic review is to analyse the effect of cytokines on EMT in gynaecological cancers and discuss their possible therapeutic implications. A search of the databases CINAHL, Cochrane, Embase, Medline, PubMed, TRIP, and Web of Science was performed using the keywords: “cytokines” AND “epithelial mesenchymal transition OR transformation” AND “gynaecological cancer”. Seventy-one articles reported that various cytokines, such as TGF-β, TNF-α, IL-6, etc., promoted EMT changes in ovarian, cervical, and endometrial cancers. The EMT changes included from epithelial to mesenchymal morphological change, downregulation of the epithelial markers E-cadherin/β-catenin, upregulation of the mesenchymal markers N-cadherin/vimentin/fibronectin, and upregulation of the EMT-transformation factors (EMT-TF) SNAI1/SNAI2/TWIST/ZEB. Cytokine-induced EMT can lead to gynaecological cancer development and metastasis and hence novel therapies targeting the cytokines or their EMT signalling pathways could possibly prevent cancer progression, reduce cancer recurrence, and prevent drug-resistance.


Introduction
The epithelial-mesenchymal transition (EMT) is a process implicated in cancer progression and metastasis, whereby epithelial cancer cells lose cellular polarity and cell-to-cell adhesions and gain metastatic and invasive properties [1]. Historically, EMT was first described in the embryonic development process, also called classical EMT [2]. However, EMT was later found to be associated with various physiological processes such as wound healing, tissue regeneration, organ fibrosis, and cancer development [2]. Inflammation is another physiological process known to promote cancer development by various molecular mechanisms, EMT being one of them [3]. EMT is said to be a pivotal point between inflammation and cancer development [4]. Inflammatory mediators such as cytokines and other soluble factors, oxidative stress, or hypoxia can promote the acquisition of EMT-like features in cancer cells. In addition, cancer cells can stimulate the secretion of cytokines and pro-inflammatory molecules that foster an inflammatory tumour microenvironment (TME) creating a self-propagating habitat for cancer cell growth [3]. The TME comprises several cell types, such as tumour cells and stroma, inflammatory and immune cells, extracellular matrix components (ECM), cancer-associated fibroblasts (CAFs), tumourassociated macrophages (TAMs), endothelial cells, epithelial cells, and mesenchymal stem cells, which are involved in the upregulation of various cytokines reported to be involved a reduction in both led to a significantly unfavourable prognosis [17] and was found to be associated with increased expression of EMT-TFs, SNAI2, and ZEB1 [18]. In addition, vimentin and SNAI1 have also been associated with the malignant phenotype of non-small cell lung cancer (NSCLC). The nuclear translocation of β-catenin was reported in a study on NSCLC [19]. In a pancreatic cancer model study by Zheng et al., SNAI1or TWISTinduced EMT was not found to be essential for the invasion and metastasis of pancreatic cancer, but they suggested that potentially targeting EMT may enhance the efficacy of chemotherapy and immunotherapy [20]. A further study using another pancreatic cancer model demonstrated that ZEB1 promoted the invasion and metastasis of cancer cells and the depletion of ZEB1 suppressed EMT-related cancer promoting changes [21]. In a study on breast cancer, SNAI1 was found to be an EMT-inducing factor as it downregulated E-cadherin, upregulated vimentin, and induced classical morphologic changes of EMT [22]. In addition to a varied expression in different cancers, the transcription factors could have antagonistic functions as well. A study on EMT in malignant melanoma reported that SNAI2 and ZEB2 transcription factors are expressed in normal melanocytes and behave as tumour-suppressor proteins, whereas TWIST1 and ZEB1 favour neoplastic transformation in melanocytes [23]. The data on the effects of EMT reversal on cancer progression/spread and drug resistance are not yet substantial, so we need robust studies to assess if reversing EMT can inhibit invasion and metastasis in cancers and counteract drug resistance.

Cytokines and EMT in Cancer Progression
Various cytokines and chemokines produced by tumour cells, CAFs, or tumourassociated immune cells in the TME can stimulate the EMT process and thereby promote cancer cell metastasis. In addition, EMT-TFs have been shown to upregulate the expression of pro-inflammatory and immunosuppressive cytokines such as TGF-β and IL-10 in cancer cells, thereby triggering tumour promoting effects on the TME.
A variety of metastasis-related chemokines (including CCL5, SDF-1, CCL2, and CCL7) and cytokines (such as IL-1, IL-6, IL-8, and TNF-α) can be released from cancer stem-like cells (CSLCs) [24]. IL-1β was reported to promote EMT in colon cancer through ZEB1 activation [24]. Another study reported that the IL-8 secreted by tumour cells undergoing EMT could then potentiate tumour progression by inducing adjacent epithelial tumour cells into EMT [25]. Onoue et al. performed a study on the effect of chemokines on EMT and demonstrated that a stromal cell-derived factor-1 (SDF-1)/CXCR4 system can facilitate lymph node metastasis in oral squamous cell carcinoma (SCC) [26]: SDF-1/CXCR4, via the activation of Phosphoinositide 3-kinase-protein kinase B (PI3K-Akt/PKB) pathway, was associated with the loss of epithelial cell morphology alongside the downregulation of epithelial markers, cytokeratin, E-cadherin, and beta-catenin and the upregulation of mesenchymal markers, vimentin, and Snail [26]. TGF-β is one of the main inducers of EMT in several biological systems [27]. Recent studies have identified two main classes of signalling pathways that are responsible for the mediation of EMT by TGFβ1: the canonical Smad signalling pathway and various non-canonical Smad-independent pathways, including the extracellular signal-regulated kinase 1/2 (ERK1/2), PI3K, c-Jun N-terminal kinase (JNK), and P38-mitogen-activated protein kinase (MAPK) pathways [28]. TGFβ1 induces SNAI1 expression in several cell types including hepatocytes, palate, and mesothelial cells. In a study on human NSCLC, IL-27 was found to induce mesenchymal morphological changes in IL-27-treated NSCLC cells, reduce epithelial markers (E-cadherin and γ-catenin) and EMT-TF Snail, and reciprocally increase mesenchymal marker vimentin predominantly through the Janus kinase-signal transducer and activator of the transcription1 (JAK/STAT1) pathway. The STAT1 pathway was implicated in the EMT changes induced by IL-27 as STAT1 inhibition reversed the EMT effects of IL-27 on IL-27-treated cells [29]. IL-6 was found to downregulate E-cadherin expression in breast cancer [30,31] and activate EMT-TF TWIST [31].
In multiple carcinomas, EMT has been demonstrated to be regulated by cytokine secretion and is possibly involved in cancer progression and metastasis including gynaecological cancers. There is a high level of drug resistance and mortality in gynaecological cancers due to metastasis and recurrence despite advances in modern diagnostic and treatment modalities. As EMT plays a crucial role in cancer metastasis, understanding the mechanism of action of the underlying EMT factors and their signalling pathways may help in reducing morbidity, mortality, and drug resistance associated with gynaecological cancers. We conducted this systematic review to summarise the evidence so far on the effect of cytokines on EMT in gynaecological cancers.

Objectives
The aim of this systematic review is to describe the effect of cytokines on EMT in gynaecological cancers and their possible therapeutic implications. This will be achieved by reporting correlations between cytokines and markers of EMT in gynaecological cancers such as morphological changes (acquisition of spindle shape), suppression of epithelial markers (E-cadherin/catenin), upregulation of mesenchymal markers (N-cadherin/vimentin/ fibronectin), and association with EMT TFs-SNAI1/SNAI2/TWIST/ZEB. The possible pathways of EMT (if mentioned in the studies) will also be documented for their potential therapeutic importance.

Protocol and Registration
A protocol for the review was devised and registered with PROSPERO (Registration No. CRD42022358266). (e) Animal studies (however, studies with both human and animal study arms have been included in the systematic review without going into any description of the animal study arm).

Information Sources
A search of the databases CINAHL, Cochrane, Embase, Medline, PubMed, TRIP, and Web of Science were performed to identify the relevant keywords contained in the titles and abstracts. A grey literature search was also performed to search for relevant conference abstracts, book chapters, leaflets, and dissertations.

Search Strategy and Selection Process
The search criteria keywords were: "cytokines" AND "epithelial mesenchymal transition OR transformation "AND "gynaecological cancer" (endometrial cancer OR ovarian cancer OR cervical cancer OR vulval cancer OR vaginal cancer). The individual search strategies are included at the end (Supplementary Materials).
Articles published in the last 22 years (from January 2000 to August 2022) in English that are indexed in the above databases were identified based on their titles and abstracts. After a perusal of the titles, duplicate studies were excluded, the abstracts for the remaining articles were reviewed, and the articles not satisfying the inclusion criteria strictly were discarded. Thereafter, full texts were obtained for the outstanding articles included so far based on their abstracts. These articles were then screened for relevance and inclusion in the systematic review for data extraction and synthesis.
The eligibility of each study was checked independently by two reviewers (I.R. and P.E.E.). The lists of included studies selected by the two reviewers were then compared and any disagreement was resolved through discussion with an independent third reviewer (L.B.M).
The search for articles for the study was performed following the new 2020 PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta Analyses Protocols) [32] and is presented in the form of a flow diagram ( Figure 1). ence abstracts, book chapters, leaflets, and dissertations.

Search Strategy and Selection Process
The search criteria keywords were: "cytokines" AND "epithelial mesenchym sition OR transformation "AND "gynaecological cancer" (endometrial cancer OR o cancer OR cervical cancer OR vulval cancer OR vaginal cancer). The individual strategies are included at the end (Supplementary Materials. Articles published in the last 22 years (from January 2000 to August 2022) in E that are indexed in the above databases were identified based on their titles and ab After a perusal of the titles, duplicate studies were excluded, the abstracts for the r ing articles were reviewed, and the articles not satisfying the inclusion criteria were discarded. Thereafter, full texts were obtained for the outstanding articles in so far based on their abstracts. These articles were then screened for relevance and sion in the systematic review for data extraction and synthesis.
The eligibility of each study was checked independently by two reviewers (I P.E.E.). The lists of included studies selected by the two reviewers were then com and any disagreement was resolved through discussion with an independent th viewer (L.B.M).
The search for articles for the study was performed following the new 2020 P guidelines (Preferred Reporting Items for Systematic Reviews and Meta Analyses cols) [32] and is presented in the form of a flow diagram ( Figure 1).

Data Extraction
The data from each study included in the review were extracted by two independent reviewers (I.R. and P.E.E.). The extracted data elements included: first author's name, publication year, study country, sample population-human samples/cell lines, cytokines/chemokines, type of gynaecological cancer studied, laboratory assays, effect of cytokine on EMT, EMT markers acquired, the possible pathways of EMT (if mentioned), and potential therapeutic implications.

Risk of Bias in Individual Studies
The studies were assessed independently for their content and methodological validity by two reviewers (I.R. and P.E.) prior to inclusion in the review. Any disagreement was resolved through discussion with an independent third reviewer (L.B.M.). The studies included were assessed for their ethical conduct and sourcing of materials.

Synthesis of Results
The studies that satisfy the inclusion criteria were divided into broad groups based on the type of gynaecological cancer: ovarian, cervical, and endometrial. No studies were found for vaginal and vulval cancers that satisfied the inclusion criteria.
After dividing into the initial three cancer groups, the findings of the studies were extracted from each study's results and discussion sections and grouped into clusters based on the EMT changes such as the epithelial-mesenchymal morphological change, downregulation of epithelial markers (E-cadherin/β-catenin/claudin/any other EMT-related epithelial marker), upregulation of mesenchymal markers (N-cadherin/vimentin/fibronectin/any other EMT-related mesenchymal marker), and alteration of EMT-TFs (SNAI1/SNAI2/TWIST/ ZEB). After dividing the EMT changes brought about by cytokines into sub-group clusters, general conclusions were drawn for each cluster in the form of narrative synthesis.

Search Results and Publication Characteristics
There were 696 articles published in the last 22 years (from January 2000 to August 2022) in English that were indexed in the above databases and were identified based on their titles and abstracts. Before screening for duplicate records, 273 studies were excluded; 423 studies were screened using their title and abstract and then 290 studies were excluded as they were not relevant to our review search criteria. Full texts were sought for the remaining 133 articles, among which 4 were conference abstracts only, so no full texts could be found and hence discarded. Next, 129 full text articles were screened for relevance and inclusion in the systematic review. A further 58 studies were excluded as they were not exactly relevant to our topic of interest, were animal studies, or did not clearly demonstrate a correlation between any cytokine and EMT changes in a gynaecological cancer. Finally, 71 articles were included in the review for data extraction and synthesis (flow diagram Figure 1). The characteristics of these studies are presented in Table 1 in the order of their year of publication and grouped with respect to the type of gynaecological cancer.

Ovarian Cancer
In this systematic review, forty-five studies on ovarian cancer have been included that satisfied the inclusion criteria and demonstrated that cytokines induce EMT changes in ovarian cancer. Various human ovarian cancer cell lines such as A2780, SKOV-3, ES-2, HEY, IGROV-1, HO-8910, and OVCAR3 or human ovarian tissues have been assessed for EMT changes. Table 2 summarises the effect of cytokines/chemokines and their mechanism of EMT in ovarian cancer. To simplify our findings, the EMT changes have been classified under the following sub-headings:

Mesenchymal morphological changes:
Mesenchymal morphological changes of ovarian cancer cells such as changes of cell shape from cobblestone (epithelial) to spindle-like narrow elongated shape (mesenchymal) have been implicated as a marker of EMT and has been reported to be induced in ovarian cancer by treatment with cytokines such as TGF-β [44,51,58,62,70,75], TNF-α [51], and chemokine CXCL12 and its receptor CXCR4 [52,53].

Downregulation of epithelial markers:
Various studies on EMT in ovarian cancer have demonstrated a downregulation of the epithelial markers E-cadherin and β-catenin induced by cytokines either in their genetic expression or protein expression or both.

Upregulation of mesenchymal markers:
Mesenchymal markers such as vimentin/fibronectin/N-cadherin have been reported to be over-expressed either at the genetic level or at the protein level on treating ovarian cancer cells or cell lines in various studies.

EMT-TF activation/suppression:
EMT TFs such as SNAI1/2, ZEB, and TWIST have been reported to be upregulated in ovarian cancer cell lines when treated with various cytokines as discussed below.

Pathways and interactions:
TGF-β has been reported to be involved in EMT in ovarian cancer by various pathways in the studies included in Table 1. The signalling pathways reported are WNT/β-catenin pathway [55] and SMAD2/3 signalling activation [40,41,75]. Left-right determination factor (LEFTY), a member of the TGF-β superfamily, was reported to be involved in the TGF-β/Smad/SNAI1 signalling for EMT induction in ovarian clear cell carcinoma cells [46]. Xu et al. [76] demonstrated that TGF-β and Epidermal growth factor (EGF) signalling pathways synergistically induce EMT and render epithelial ovarian cancer cells a more invasive phenotype. Matrix metalloproteinases (MMP2/9) are involved with cell motility and are closely related to EMT and were reported to be upregulated by TGF-β in various studies [39][40][41]56,76]. A study by Ren et al. [36] demonstrated that TGF-β treatment decreased miRNA-200 expression. MicroRNAs (miRNAs) are small non-coding single-stranded RNAs that control gene expression by targeting mRNA translation. MiRNA-200 has been reported to have a suppressive effect on EMT by inhibiting transcriptional repressors ZEB1 and ZEB2. The authors reported that miRNA-200 inhibited EMT by downregulating the sex-determining region Y-box 4 (SOX4), which is an upstream factor for EMT. They also demonstrated that TGF-β treatment induced EMT by decreasing miRNA-200 expression. The TGF-β/ZEB/miR-200 signalling pathway, an autocrine regulatory network, has been reported by Gregory et al. [104] to control the plasticity between the epithelial and mesenchymal states of the cells. TGF-β signalling activated ZEB1/2, which in turn induced EMT by repressing epithelial genes. Furthermore, miRNA-200 was noted to suppress ZEB1/2 and promote epithelial differentiation. ZEB1/2 knockdown enhances miRNA-200 expression and TGF-β signalling is a target of miRNA-200. However, exogenous TGF-β administration can inhibit miRNA-200. Hence, there is an interconnection between TGFβ, miRNA-200, and ZEB that is essential in regulating the epithelial and mesenchymal reversible states [105].
Similar to TGF-β, IL-6 has been reported to be involved in EMT by various pathways. Wang et al. [47], Ma et.al. [48], and Colomiere et al. [77] reported that IL-6 mediated the EMT in OVCAR3 cells via the JAK2/STAT3 pathway. IL-6 secretion was reported to be induced by CAFs in the TME [47] and the IL-6/IL-6R/STAT3 signalling was reported to be induced by CD146 by Ma et.al. [48] and by the epidermal growth factor (EGF) by Colomiere et.al. [77].
The IL-8 and IL-8 receptors CXCR1 and CXCR2 were reported by Wen et al. [37] and Yen et al. [66] to induce EMT changes in ovarian cancer cells potentially by the Wnt/β-catenin pathway, similarly to TGF-β.
IL-17 treatment possibly induced EMT in ovarian cancer cell lines via the expression of metastasis-associated genes-1 (MTA1) and targeting the IL-17/MTA-1 axis could be used as a treatment for ovarian cancer [38].
GDF8, which belongs to the TGF-β superfamily, was reported by Zhou et al. [63] to promote EMT in ovarian cancer cells via Activin such as the kinase 4 and 5 (ALK4 and 5) pathways.
Chemokines have also been reported to activate various EMT signalling pathways. The NF-κB signalling pathway has been implicated in CCL5-induced EMT changes in ovarian non-cancer stem-cell-like cells [67] and the AKT/ERK pathway was reported to be activated by CXCR7 and its ligand CCL19 in studies by Yu et al. [68] and Cheng et al. [69].
Therapeutic possibilities: A number of molecules or factors have been implicated in the studies included in this systematic review to either potentiate or abrogate cytokine-mediated EMT signalling and hence targeting them could open new therapeutic avenues.
In a study by Cheng et al. [74], TGF-β-induced EMT changes were demonstrated in serous borderline ovarian tumour (SBOT) cells, indicating that TGF-β-induced EMT is possibly involved in the progression from non-invasive SBOT to invasive low grade ovarian cancer (LGC) and that targeting the TGF-β signalling pathway could prevent the progression from borderline ovarian tumour to ovarian cancer. However, Sicard et al. [35] reported that TGF-β-induced mesenchymal changes were limited to the chemo-sensitive ovarian cancer cells and Ameri et al. [39] observed that TGF-β-mediated EMT was more prominent in epithelial-like ovarian cancer cell lines than invasive ovarian cancer cell lines. Therefore, strategic targeting of TGF-β signalling in non-invasive or chemo-sensitive ovarian cancers may have an increased therapeutic benefit. TGF-β has been demonstrated to induce EMT changes in ovarian cancer cells by inducing DNA methyltransferases (DNMT) that are involved in DNA methylation [71]. DNA methylation has been implicated in suppressing various EMT genes. Treatment with DNMT inhibitor (SGI-110) prevented TGF-β induced EMT changes. Hence, targeting DNMT may reverse the EMT changes or EMT gene suppressions caused by DNA methylation in ovarian cancer [71].
BMP9, a member of the TGF-β superfamily, has been demonstrated by Wang et al. [42] to promote EMT changes in a dose-dependent manner; the authors have also commented that BMP9-induced EMT may be partially responsible for BMP9-induced Cisplatin chemoresistance in ovarian cancer. This raises the possibility that BMP9 could be a novel therapeutic target to improve cisplatin sensitivity in chemo-resistant patients [42].
In the case of IL-6-induced EMT signalling in ovarian cancer, receptor interacting protein serine/threonine kinase 4 (RIPK4) has been reported to be of significance. RIPK4 is a key member of the group of Receptor Interacting Proteins (RIPs) and is aberrantly expressed in multiple cancer types [34]. Silencing RIPK4 significantly downregulated IL-6-mediated EMT changes [34] and, hence, dual targeting of RIPK4/IL6 could have a therapeutic advantage. Wang et al. reported that IL-6-induced EMT can enhance paclitaxel resistance in ovarian cancer cells [47] and, hence, it is possible that reversing the IL-6mediated EMT may reverse drug resistance. However, substantial studies are needed to verify this possibility. Table 2. EMT changes associated with cytokines/chemokines in ovarian cancer.

Cervical Cancer
Eighteen studies on cervical cancer have been included in this systematic review that satisfied the inclusion criteria. Various EMT changes have been demonstrated by the treatment of cervical cancer cell lines such as C33a, Hce1, HeLa, CaSki, SiHa, etc., or human cervical cancer tissue with different cytokines/chemokines by immunoblot, immunofluorescence, or qRT-PCR. Table 3 summarises the EMT changes associated with cytokines/chemokines in cervical cancer. The EMT changes in cervical cancer have been classified under sub-headings, as follows:

Mesenchymal morphological changes:
Morphological change of cervical cancer cells-change of cell shape from cobblestone (epithelial) to spindle-like narrow elongated shape (mesenchymal)-has been implicated as a marker of EMT and reported to be induced in cervical cancer cells by treatment with TGF-β [78,86,[88][89][90]92,95] as well as TNF-α [86].
Cytokine with anti-tumour effect: IL-9, a T-helper 9 cytokine, reduced the expression of N-cadherin and vimentin in cervical cancer cells and increased expression of E-cadherin. Therefore, IL-9-based therapy has the potential to prevent progression and metastasis in cervical cancer [79].
Miao et al. demonstrated that STAT3 is involved in IL-6-induced EMT changes in HeLa and C33A human cervical cancer cells and STAT3 silencing led to a reversal of IL-6-induced EMT changes [94].
Therapeutic possibilities: As in ovarian cancer, different molecules have been implicated in cervical cancer studies to promote or abrogate specific cytokines or their signalling pathways involved in EMT. TGF-β-mediated EMT changes could be blocked by Hesperetin, a flavonoid in citrus fruits [78]; Epigallocatechin-3-gallate (EGCG), a polyphenolic compound found in green tea [80] and Chalcone L1, which is a natural antioxidant and anti-inflammatory polyphenol sourced from plants [81]. Therefore, Hesperetin/EGCG/Chalcone L1 may have therapeutic benefits as anti-cancer agents for the treatment of cervical cancer. In a study by Li et al. [82] cadherin CDH 20 (belonging to a superfamily of cell-to-cell adhesion molecules) interacted with β-catenin and suppressed TGF-β-mediated EMT in cervical cancer cell lines. This suggests that CDH 20 may act as a tumour suppressor that can inhibit cervical cancer cell migration and invasion and hence may have therapeutic potential. Wu et al. [85] demonstrated that TGF-β1 induced EMT in cervical cancer cells in both Human papillomavirus (HPV)-positive and negative cervical cancer cells. Therefore, TGF-β1 could be used for targeted therapy for cervical cancer irrespective of HPV status. The knockdown of some molecules involved in TGF-β/EMT signalling could halt this EMT pathway and could be used for cancer treatment; such molecules could be CD36, a membrane glycoprotein present on various epithelial cells [83]; RhoE, a RNA/DNA helicase [84]; FAD104, a fibronectin type III domain-containing protein (FNDC) [87]; p68, a type of RNA helicase [89]; and Sine oculis homeobox homolog 1 (SIX1), a transcription factor associated with development but rarely expressed in adults [91]. Hence, targeting CD36, RhoE, FAD104, p68, or SIX alongside TGF-β signalling inhibition could be new therapeutic avenues for cervical cancer treatment.
Zhang et al. [93] reported that the expression of astrocyte-elevated gene-1 (AEG), a multifunctional oncoprotein, was increased by treatment with chemokine CCL20/CCR6. AEG knockdown resulted in abrogation of the EMT changes and disruption of the ERK1/2-Akt signalling induced by CCL20. This implies that AEG is an important component of the CCL20/CCR6-Erk1/2-Akt-EMT pathway and could be a novel targeted therapy for cervical cancer. Epithelial downregulation: E-cadherin [94].

Endometrial Cancer/Uterine Cancer
In this systematic review, nine studies on endometrial cancer have been included (one overlapping with ovarian cancer). Various human endometrial cancer cell lines such as HEC-1A, HEC 1B, and Ishikawa cells or human endometrial cancer tissue have been examined for EMT on treatment with various cytokines and chemokines by multiple methods such as immunoblot, immunofluorescence, or quantitative reverse transcriptase PCR (qRT-PCR). Table 4 documents the EMT changes associated with cytokines or chemokines in endometrial cancer. To simplify our findings, the EMT changes in endometrial cancer have been classified under the following sub-headings:
EMT-TF activation/suppression: SNAI1, SNAI2, TWIST, and ZEB EMT TFs have been reported to be involved in EMT in endometrial cancer cell lines.

Possible pathways and interactions:
Smad2/3/TGF-β signalling has been reported to be involved in EMT in endometrial cancer cell lines [80]. Chen et al. [96] reported that TGF-β1 possibly induces EMT by the Smad3/TWIST signalling pathway in endometrial cancer cells. Li et al. reported that AMF induces EMT in endometrial cancer via the transforming growth factor β receptor 1 (TGFBR1)/ERK/MAPK pathway [103].

Therapeutic possibilities:
Chen et al. [96] reported that Isoliqueritigenin (ISL), a flavonoid derived from liquorice and bean sprouts, may be able to reverse TGF-β-induced EMT changes in endometrial cancer cell lines, causing it to be a potential candidate for targeting TGF-β-induced signalling in endometrial cancer. The treatment of an endometrial cancer cell line with fluorene-9bisphenol (BHPF), a derivative of bisphenol A, in the study by Wang et al. [98] inhibited the TGF-β1-induced expression of EMT markers, indirectly demonstrating the effect of TGF-β1 on EMT and indicating that BHPF could be used as a possible novel therapy for endometrial cancer.

Risk of Bias in Individual Studies
Each study was evaluated for their methodological integrity, ethical sourcing of materials, and conduct of study before including in the review. The studies that did not clearly state the active compound being investigated or clearly demonstrate in their methodology and results a specific EMT change as a result of treatment with a specific cytokine in gynaecological cancers were excluded from the review. It was not possible to evaluate the quality and methodological soundness of conference abstracts and, hence, they were not included in the study.

Discussion
Various cytokines and chemokines have been demonstrated to be involved in promoting EMT in gynaecological cancers, such as TGF-β, IL-6, IL-8, and TNF-α. These soluble mediators along with various growth factors such as EGF, fibroblast growth factor (FGF), hepatocyte growth factor (HGF), platelet-derived growth factor (PDGF), tumour growth factor (TGF)-β, and vascular endothelial growth factor (VEGF) along with immune cells in the TME such as the immune infiltrating macrophages, CAFs, neutrophils, and platelets can promote EMT in primary tumours. They can transform epithelial cells to mesenchymal cells by upregulating a variety of EMT transcriptional factors, such as SNAI1/SNAI2/ZEB/TWIST that are repressors of epithelial gene such as E-cadherin and activators of mesenchymal genes such as N-cadherin/fibronectin/vimentin, α-smooth muscle actin, etc. Mesenchymal cells invade the surrounding stroma and eventually enter the systemic circulation, reach distant sites and undergo mesenchymal-epithelial transformation (MET), which is essential for the outgrowth of metastases.
Similar to studies on oral SCC [26], NSCLC [29], and breast cancer [30,31], cytokines or chemokines such as interleukins and SDF-1/CXCR4 have been involved in the reduction in the expression of epithelial markers such as E-cadherin in ovarian, endometrial, and cervical cancers and the upregulation of mesenchymal markers such as N-cadherin and/or vimentin. These cytokines regulate EMT via various mechanisms by activating EMT-TFs such as TWIST/SNAI1/SNAI2/ZEB and signalling pathways. The predominantly reported EMT signalling pathways among a multitude of pathways and interactions mentioned above involve WNT/β-catenin activation [55] and SMAD2/3 activation by TGF-β [40,41,75] and STAT3 activation by IL-6 [47,48,77]. The roles of cytokines are complicated and they can interact among themselves and with other substances in an inflammatory tumour microenvironment and create a complex permissive milieu of signalling that regulate EMT changes and promote cancer progression and metastasis and also possibly induce drug resistance and allow for the recurrence of cancer.

Strengths and Limitations of Our Review
The strength of our review is that it presents a comprehensive review of all the English language studies in the last 22 years demonstrating the effect of cytokines/chemokines on EMT in all gynaecological cancers under one platform. However, the use of a specified time range and English language for our search criteria may have introduced a time-period bias and a language bias, respectively. We have performed a grey literature search to reduce the risk of selection bias. The PRISMA 2020 guideline [32] has been used for presenting screened, excluded, and analysed articles in the form of a flow diagram to reduce publication and selection bias. All the studies included used either human tissue or gynaecological cancer cell lines, mentioned the source of all tissues, cell lines, and reagents used and described in detail the tests involved such as immunoblot/immunofluorescence/qRT-PCR.
The limitations are that the articles included are not uniform in their cell type or tissue and modes of experiments. Some studies have used human tissue, while others have used cancer cell lines and some both. Among the cell lines, different sources have been used across the studies included in the systematic review. Due to the different sources/methods/treatment conditions used, it was not possible to synthesise the results using a single statistical test and hence no statistical tests have been used.

Further Recommendations for Research
We did not find any studies satisfying the inclusion criteria for our systematic review on vulval and vaginal cancers. These cancers are closely anatomically related to cervical cancer and cytokine-mediated EMT has been shown to have a role in cervical cancer progression and metastasis. Hence, further studies are required to investigate the role of cytokines and other soluble factors on EMT in these two cancers. The role of EMT in inducing drug resistance in gynaecological cancers needs further investigation. Additionally, the possibility of halting cancer progression or reversing drug resistance by targeting specific cytokines or their downstream effector molecules and/or EMT markers and EMT-TFs needs extensive exploration.

Conclusions
Various studies have been included in this systematic review to document the effect of cytokines on EMT in gynaecological cancers, using different cell lines or human tissue from multiple sources. To the best of our knowledge, this study is the only systematic review on the effect of cytokines on EMT in gynaecological cancers. The review summarises all the relevant English language articles on this subject in the last 22 years. However, the studies are not yet exhaustive and further investigations are warranted to explore the roles of cytokines on EMT in gynaecological cancers in depth. EMT is an essential part of cancer progression and, hence, it is possible that targeted therapy affecting these cytokines themselves or either upstream or downstream EMT signalling pathways could be of therapeutic benefit in cancer treatment. Defining the exact role of the cytokines involved and their signalling pathways inducing EMT in a specific type of gynaecological cancer and developing novel targeted therapies could potentially halt cancer progression or metastasis, reduce the risk of cancer recurrence, reverse or prevent drug-resistance in these cancers, and could also supplement available chemotherapy and other therapies appropriately.
Supplementary Materials: The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/cells12030416/s1, File S1: Detailed search criteria for the systematic review according to PRISMA guidelines.
Author Contributions: I.R. and P.E.E. conceptualised this review, developed the protocol, selected the articles for full-text review, performed data extraction and synthesis, and wrote the original manuscript. L.B.M. reviewed the manuscript, made suggestions regarding inclusion and exclusion of articles, and helped in editing the review. A.M. also reviewed the manuscript and provided valuable input. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.

Institutional Review Board Statement:
In this systematic review, we used only previously published data. As no unpublished data was used, we did not seek ethics committee approval. The review layout is according to the PRISMA 2020 guidelines for systematic reviews [32].

Informed Consent Statement:
No special software or code has been used. Search criteria for articles have been mentioned in the review.
Data Availability Statement: Not applicable, as review article data were derived from other studies that have been mentioned using references.