Venous Thrombogenesis and Cervical Cancer: Plasma MicroRNAs as Prognostic Indicators of Tumor Behavior

Abstract

Cervical cancer (CC) is the fourth most common cancer among women globally, with venous thromboembolism (VTE) representing a life-threatening complication. Cancer-associated thrombosis (CAT) arises from tumor-driven activation of hemostasis, worsening prognosis. Recently, circulating microRNAs (miRNAs) have emerged as potential biomarkers for both CAT and cervical tumorigenesis. Thus, this study aimed to assess the implications of five miRNAs—miR-20a-5p, -23a-3p, -125b-5p, -145-5p, and -616-3p—in CC-related VTE context. These miRNAs were quantified by RT-qPCR in plasma from 69 CC patients before treatment. Briefly, VTE occurred in nine patients, decreasing overall survival (OS) [log-rank test, p = 0.005; hazard ratio (HR) = 4.78; 95% confidence interval (CI), 1.42–16.05]. Lower miR-20a-5p levels predicted VTE (ꭓ² test, p = 0.027) and, in subgroup analyses, they were linked to cervical squamous cell carcinoma (CSCC) and older age (ꭓ² test, p = 0.003 and p = 0.043, respectively). In VTE patients, miR-145-5p downregulation was associated with improved OS (log-rank test, p = 0.018), an effect also observed in the adenocarcinoma (ADC) subgroup (log-rank test, p = 0.039). The remaining miRNAs showed subtype-specific links to clinicopathological features and survival. These findings highlight the potential value of circulating miRNAs in thrombotic risk and prognosis assessment in CC.

1. Introduction

Cervical cancer (CC) ranks as the fourth most diagnosed cancer and the fourth leading cause of cancer-related death among women worldwide, with approximately 660,000 new cases and 350,000 deaths in 2022 [1,2]. The disease mainly results from persistent infection with high-risk human papillomavirus (HPV), particularly types 16 and 18 [3,4]. Histologically, CC includes two main subtypes: squamous cell carcinoma (CSCC), which is the most common but with a decreasing incidence due to effective prevention through HPV vaccination and early detection, and adenocarcinoma (ADC) [5,6,7].

Despite the decrease in CC incidence and mortality, especially in high-income countries with organized vaccination and screening programs [8,9], CC patients face additional challenges, notably cancer-associated thrombosis (CAT) [10]. This condition mainly involves venous thromboembolism (VTE) events, which occur at a significantly higher rate among cancer patients (up to seven-fold compared to the general population), and it is recognized as the second leading cause of death in this subpopulation [11,12,13,14]. In CC, the average CAT incidence ranges between 9% and 12% and is linked to poor survival [15,16]. This increased pro-thrombotic risk reflects the ability of CC cells to release pro-coagulant, pro-angiogenic and inflammatory factors that enhance tumor fitness and facilitate disease progression. Combined with the effects of tumor treatments, these changes sustain a pro-thrombotic state, ultimately contributing to the clinical manifestation of CAT [17,18,19].

Regarding the disease pathogenesis, thrombosis is defined as the formation of an abnormal blood clot (thrombus) within a blood vessel, which can partially or completely block blood flow [20]. Under normal physiological conditions, hemostasis maintains vascular integrity after injury, allowing normal blood flow through vasoconstriction, primary hemostasis (platelet adhesion and aggregation) and secondary hemostasis (activation of the coagulation cascade to form a stable fibrin clot) [21]. The coagulation cascade proceeds through the extrinsic pathway, triggered by tissue factor (TF) exposure, and the intrinsic pathway, activated by contact with negatively charged surfaces [22,23,24]. Both converge on thrombin generation, which converts fibrinogen into fibrin, stabilizing the platelet plug—common coagulation cascade [25,26]. Once vascular repair begins, fibrinolysis removes the clot through plasmin-mediated degradation of fibrin [27]. Several inhibitory mechanisms operate to prevent excessive coagulation [28]. One of them is tissue factor pathway inhibitor 1 (TFPI1 or TFPI), which binds and inhibits the TF-activated coagulation factor VII (FVIIa) complex in an activated coagulation factor X (FXa)-dependent manner, thereby controlling initiation of the extrinsic coagulation pathway and limiting excessive thrombin and fibrin generation [29]. Its homolog, tissue factor pathway inhibitor 2 (TFPI2), is involved in the regulation of intrinsic clotting by targeting plasma kallikrein (pKLK) and activated coagulation factor XI (FXIa). However, it also modulates fibrinolysis by targeting plasmin [29,30,31].

MicroRNAs (miRNAs) are small non-coding RNAs (ncRNAs) that regulate gene expression post-transcriptionally, influencing multiple biological pathways, including hemostasis [32]. A defining feature of miRNA activity is that each miRNA can regulate the expression of multiple messenger RNAs (mRNAs), and, conversely, a single transcript can be regulated by different miRNAs [33,34,35]. Beyond their intracellular roles, these ncRNAs can be found circulating in body fluids, including plasma and serum. Their extracellular presence is largely attributed to their remarkable stability and resistance to degradation by endogenous RNases, making them valuable biomarkers [36,37]. Focusing on hemostasis, miRNAs are key regulators of the coagulation system by directly targeting hemostatic genes such as fibrinogen (coagulation factor I), coagulation factor 11 (F11) and coagulation factor 3 (F3) (encoding TF protein), as well as anticoagulants such as TFPI, antithrombin and plasminogen activator inhibitor-1 (PAI-1) [36,38,39].

Given the central role of the TF pathway in CAT onset, miRNAs targeting F3, TFPI and TFPI2 may provide prognostic value among CC patients. Previously, our research group conducted a comprehensive review of these miRNAs [29]. More recently, an exhaustive review of the existing literature on miRNAs involved in cervical carcinogenesis was conducted. Considering both assessments, miR-20a-5p, -23a-3p, -125b-5p, -145-5p and -616-3p emerge as potential multipurpose biomarkers for early CC diagnosis, prognostication, CAT risk stratification and therapeutic guidance in CC patients. The present study aimed to evaluate the implications of these five miRNAs in a cohort of 69 CC patients from the Northern Region of Portugal, considering the bidirectional interplay between tumor progression and thrombosis. More specifically, the impact of these circulating miRNAs on the development of venous thrombotic events and patient prognosis, independently of VTE, was examined. To our knowledge, this is the first study to explore the potential role of miRNAs in linking TF pathway to both CAT occurrence and CC behavior, highlighting their prognostic relevance in CC patients.

2. Results

2.1. Correlation Between Baseline miRNA Expression

Except for miR-616-3p, which only showed moderate positive correlations with miR-20a-5p (ρ = 0.555), all pairwise correlations between plasma miRNA expression levels were significant and positive, with coefficients ≥ 0.500 (Figure 1 and

Table 1. Correlation analysis between miRNA expression levels.

miRNA		miR-20a-5p	miR-23a-3p	miR-125b-5p	miR-145-5p	miR-616-3p
miR-20a-5p	Coefficient test	-	0.659 *	0.598	0.662 *	0.555
	p-value		<0.001	<0.001	<0.001	<0.001
miR-23a-3p	Coefficient test	0.659 *	-	0.546	0.700 *	0.305
	p-value	<0.001		<0.001	<0.001	<0.001
miR-125b-5p	Coefficient test	0.598	0.546	-	0.660	0.332
	p-value	<0.001	<0.001		<0.001	<0.001
miR-145-5p	Coefficient test	0.662 *	0.700 *	0.660	-	0.434
	p-value	<0.001	<0.001	<0.001		<0.001
miR-616-3p	Coefficient test	0.555	0.305	0.332	0.434	-
	p-value	<0.001	<0.001	<0.001	<0.001

Spearman’s rank correlation coefficient test was used when at least one of the miRNAs did not show normal distribution. * Pearson’s rank correlation coefficient test was calculated when both miRNAs followed a normal distribution. According to the literature, the strength of the correlations (Spearman’s ρ and Pearson’s r ≥ 0.500 and p < 0.05) was classified as follows: moderate (Spearman’s ρ and Pearson’s r < 0.700), strong (0.700 ≤ Spearman’s ρ and Pearson’s r < 0.900), and very strong (Spearman’s ρ and Pearson’s r ≥ 0.900) [40,41,42]. Abbreviation: miRNA, microRNA.

). The strongest correlation was observed among miR-23a-3p and miR-145-5p (Pearson’s r = 0.700).

2.2. MiRNA Expression Levels and Patients’ Characteristics

The distribution of miRNA expression varied significantly across patient characteristics, except for miR-616-3p (

Table 2. Pattern of miRNA expression according to patients’ characteristics.

miRNA	Age (Years)		Histology			FIGO Stage
	<52	≥52	CSCC	ADC	Other	<IIB	≥IIB
	N (%)	N (%)	N (%)	N (%)	N (%)	N (%)	N (%)
miR-20a-5p
Low	7 (21.2%)	15 (48.4%)	31 (60.8%)	2 (20%)	0 (0%)	6 (66.7%)	27 (49.1%)
High	26 (78.8%)	16 (51.6%)	20 (39.2%)	8 (80%)	3 (100%)	3 (33.3%)	28 (50.9%)
p-value	0.043 *		0.003			0.332
miR-23a-3p
Low	13 (39.4%)	20 (58.8%)	28 (53.8%)	4 (33.3%)	1 (33.3%)	9 (100%)	36 (62.1%)
High	20 (60.6%)	14 (41.2%)	24 (46.2%)	8 (66.7%)	2 (66.7%)	0 (0%)	22 (37.9%)
p-value	0.178		0.194			0.025 **
miR-125b-5p
Low	12 (36.4%)	21 (61.8%)	21 (40.4%)	1 (8.3%)	0 (0%)	5 (55.6%)	28 (48.3%)
High	21 (63.6%)	13 (38.2%)	31 (59.6%)	11 (91.7%)	3 (100%)	4 (44.4%)	30 (51.7%)
p-value	0.067		0.018 *			0.687
miR-145-5p
Low	13 (39.4%)	20 (58.8%)	29 (55.8%)	4 (33.3%)	0 (0%)	8 (88.9%)	25 (43.1%)
High	20 (60.6%)	14 (41.2%)	23 (44.2%)	8 (66.7%)	3 (100%)	1 (11.1%)	33 (56.9%)
p-value	0.178		0.027			0.011
miR-616-3p
Low	14 (43.8%)	17 (54.8%)	27 (54%)	4 (40%)	0 (0%)	5 (62.5%)	26 (47.3%)
High	18 (56.2%)	14 (45.2%)	23 (46%)	6 (60%)	3 (100%)	3 (37.5%)	29 (52.7%)
p-value	0.530		0.070			0.425

All p-values correspond to comparisons between low and high miRNA expression levels (expression profile 1. See Section 4.7. Statistical Analysis). p-values marked with * refer to comparisons between low vs. medium–high expression levels, while those marked with ** refer to comparisons between low–medium vs. high expression levels. Abbreviations: miRNAs, microRNAs; CSCC, Cervical Squamous Cell Carcinoma; ADC, Cervical Adenocarcinoma; FIGO, International Federation of Gynecology and Obstetrics.

). Reduced miR-20a-5p levels were more common in older patients (≥52 years) and those with CSCC tumors (ꭓ² test, p = 0.043 and p = 0.003, respectively). Patients at early International Federation of Gynecology and Obstetrics (FIGO) stages (<IIB stages) more commonly exhibited low-to-medium expression levels of miR-23a-3p (ꭓ² test, p = 0.025). Regarding miR-125b-5p, patients with ADC tumors tended to display higher miRNA expression levels (ꭓ² test, p = 0.018). As for miR-145-5p, its lower expression was more common among patients with CSCC and those at early disease stages (ꭓ² test, p = 0.027 and p = 0.011, respectively).

2.3. MiRNA Expression Levels and CC-Related VTE Incidence

The association of the evaluated miRNAs with VTE status in CC patients [without VTE versus (vs.) with VTE after CC diagnosis] was initially assessed using the Mann–Whitney U test or an unpaired t-test, depending on data normality. There were no associations between the miRNAs and VTE development (Figure 2). However, as a subsequent analysis, a ꭓ² test was performed to explore whether the expression pattern of the miRNAs vary according to VTE status. This analysis revealed that VTE-free patients more frequently presented an elevated expression of miR-20a-5p (expression profile 1—see Section 4.7. Statistical Analysis) compared to those who developed the condition after CC diagnosis (72.0% and 22.2%, respectively; ꭓ² test, p = 0.012). No further significant differences were observed for the remaining miRNAs (ꭓ² test, p > 0.05). Consistently, logistic regression analysis corroborated the association between miR-20a-5p and VTE status. Specifically, those with high levels of this miRNA (expression profile 1) exhibited a reduced likelihood of VTE [odds ratio (OR) = 0.12; 95% confidence interval (CI), 0.02–0.71], indicating a potential protective role of this miRNA. No further significant associations were detected (p > 0.05).

2.4. VTE Status and Patients’ Prognosis

The occurrence of venous thrombotic events among CC patients was shown to negatively impact their survival (log-rank test, p = 0.005; Figure 3). Considering only the patients with available miRNA expression data (N = 66), those who developed the condition after cancer diagnosis (N = 9) had a markedly shorter mean overall survival (OS) (30.2 ± 4.3 months) compared to their counterparts (49.8 ± 2.0 months). Indeed, developing VTE after CC diagnosis was linked to a nearly fivefold increased risk of death within 10 years [hazard ratio (HR) = 4.78; 95% CI, 1.42–16.05; p = 0.011].

2.5. MiRNA Expression Levels and Patients’ Prognosis

Considering the entire cohort (N = 66), no significant association was exhibited between baseline expression levels of any of the miRNAs studied and patients’ OS. In contrast, significant associations were identified in the subgroup analyses. Namely, among patients with ADC tumors, a higher OS was observed for those with reduced levels of miR-23a-3p, miR-145-5p and miR-616-3p (log-rank test, p < 0.05; Figure 4a, Figure 4b and Figure 4c, respectively).

In the stratified analysis according to VTE status, in the subgroup of patients with clinical evidence of thrombosis occurrence, those with low-to-medium plasma levels of miR-145-5p had a higher OS (log-rank test, p = 0.018; Figure 5). As for the subgroup analyses considering patients’ age and FIGO stage, no statistically significant association between miRNA baseline expression levels and patient OS was observed (log-rank test or Tarone–Ware test, p > 0.05).

2.6. In Silico Analysis

Given the impact of miR-23a-3p, miR-145-5p, and miR-616-3p on patient prognosis, the shared biological implications of these miRNAs were explored through an in silico analysis. First, validated targets of each miRNA (247, 230 and 53 targets for miR-23a-3p, miR-145-5p, and miR-616-3p, respectively) were identified using miRTargetLink 2.0 (Appendix A 

Table A1. Validated targets of miR-23a-3p, miR-145-5p and miR-616-3p according to MirTargetLink 2.0.

Validated Targets According to MirTargetLink 2.0 *
miR-23a-3p			miR-145-5p			miR-616-3p
APAF1	DNAJC21	PTPDC1	ABCC1	NAIP	IVNS1ABP	PON1
ATAT1	DOCK4	PTPN14	ABHD17C	NANOG	KIF21A	TFPI2
CDH1	DSTN	QSER1	ABRACL	NDRG2	LMNB2	BBOX1
CHUK	DYNC2LI1	RNF111	ADAM17	NDUFA4	MAP1LC3B	C5orf46
CXCL12	DYNLL2	RNF38	ADD3	NEDD9	MAP2K4	CAPRIN2
CXCL8	EGLN3	RNMT	AKR1B10	NFATC1	MAP3K11	CARNS1
FANCG	EN2	RPL7L1	ALDH3A1	NIPSNAP1	MAP3K3	CCSAP
FAS	ERBIN	RPP14	ALPPL2	NRAS	MAP4K2	CDK4
FOXA1	ETNK1	RPRD2	ANGPT2	NUDT1	MUC19	CEP350
FOXO3	FAM222B	RPS4X	AP1G1	PAK4	MUC4	CNNM2
FZD5	FAM91A1	RPS5	APH1A	PARP8	NDUFS2	COIL
G6PC	FASTKD1	RPSAP58	ARF6	PIGF	NR1D2	CREBZF
GJA1	FASTKD5	S100A7A	ARL6IP5	PODXL	NUFIP2	CSGALNACT1
GLS	FAU	SCAMP2	BNIP3	POU5F1	NUP43	CXCL5
HES1	FBN2	SDHD	BRAF	PPP3CA	ORC4	DCTPP1
HMGB2	FKBP4	SEMA6D	C11orf65	PTP4A2	P4HA1	DLGAP3
HMGN2	FLNA	SEPT2	CAMK1D	PXN	PADI1	FSHB
HNF1B	FNIP1	SERINC3	CBFB	ROBO2	PANK1	HLA-DOB
HOXB4	FOXM1	SESN2	CCDC43	ROCK1	PDGFD	IGFBP5
HSP90AA1	FOXR2	SLC27A4	CD28	RPA1	PHACTR2	INTS6
IL6R	FSD1L	SLC35G3	CD40	RPS6KB1	PLAGL2	KMT2A
IRF1	FUT4	SLC6A15	CD44	RREB1	PLEKHM1	LGALS3BP
KLF3	GGA3	SLC6A6	CDH2	RTKN	PNMA3	LPAR2
LAMP1	GMPS	SNTG1	CDK4	SENP1	PRDM2	MAL2
LDHA	GNAI3	SOCS6	CDK6	SERINC5	PSAT1	MC2R
LDHB	GNB2	SOD2	CDKN1A	SERPINE1	RAB3IP	MED17
LPAR1	GRK2	SPTY2D1	CEP19	SET	RBM18	MIEF1
LRP5	GRTP1	SSR1	CFTR	SMAD2	REL	MIER3
MEF2C	GTF3C4	SSRP1	CLINT1	SMAD3	RPS6KA3	MRO
MT2A	HAPLN1	STAMBPL1	COL5A1	SOCS7	RRAGC	MTHFD1
MYH1	HCFC1	STARD7	CRNDE	SOX2	SAMD5	PAFAH1B1
MYH2	HIST1H3B	STS	CTGF	SOX9	SESN2	PRKAG1
MYH4	HIVEP1	STT3B	CTNND1	SP1	SLC16A10	PRPF4B
NEK6	HMGCS1	SWAP70	DDC	SP7	SLC16A5	PSMA2
POU4F2	HNRNPUL1	TBL2	DDX17	SPTBN1	SLC22A9	PTMA
PPARGC1A	HRG	TCF25	DDX6	SPTLC1	SLC26A2	QSER1
PPP2R5E	HS2ST1	THUMPD3	DFFA	SRGAP1	SMAD4	RBM26
PTEN	IMPDH2	TLR6	DTD1	STAT1	SMAD5	RBM43
PTPN11	ITPKC	TMED7	E2F3	SWAP70	SMIM17	SERTAD2
RGS5	JMJD1C	TMEM170A	EGFR	TGFB2	SNTB1	SFMBT2
SMAD3	KDM3B	TNFAIP3	EIF4E	TGFBR2	SNX24	SIK1
SMAD5	KIAA1210	TNFAIP8	EPAS1	TIRAP	SOX11	SLFN12
SPRY2	KIAA1551	TNPO1	ERG	TMEM9B	SRPX2	SLFN12L
ST7L	KIF20A	TNRC6A	ETS1	TNFSF13	TGFBI	SMAD4
STAT3	KIF22	TOPBP1	F11R	TPM3	THSD7A	TMEM59
TERF2	KLF10	TOR1AIP1	FAM3C	TPRG1	TMOD3	TOR4A
TMEM64	KLF12	TPM3	FAM45A	TSPAN6	TNR	TRMT112
TOP1	LEMD2	TRIM59	FLI1	TUG1	UBR7	USP37
TSC1	LMAN2	TRIM63	FSCN1	VEGFA	UBXN2A	XIAP
XIAP	LMNB1	TRRAP	FXN	VPS51	UNC5D	ZNF347
ABCD1	MAP1B	TSNAX	GMFB	YES1	UTP15	ZNF35
ACSS3	MC2R	TXNIP	GOLM1	AAED1	VGLL4	ZNF460
ACTN4	MCFD2	TXNL4A	HDAC11	ACTB	WASHC2C	ZNF850
ADAM17	MCU	UBALD1	HDAC2	AGPS	WSB1	PON1
ADAM28	MGAT5	UQCRFS1	HLTF	AGTRAP	ZBTB25
ADAM9	MMGT1	USP34	HMGA2	ALG9	ZFAND3
ADD3	MPP6	USP5	IFNB1	ANKRD28	ZFYVE9
ARL6IP1	MRPS34	VAV3	IGF1R	AQR	ZNF100
ASNS	MTAP	VCAM1	ILK	BCLAF3	ZNF426
ATP5A1	MYC	WEE1	IRS1	BTG1	ZNF445
ATXN7L3B	MYH10	ZBTB10	IRS2	C1orf27	ZNF451
BAZ2A	NACC1	ZCCHC2	ITGB8	CCDC80	ZNF660
BCAP29	NDUFV3	ZIK1	JADE1	CCDC85C	ZNF678
BRWD1	NLGN4X	ZNF117	KLF4	CLSTN2	ZNF772
BSDC1	NLK	ZNF138	KLF5	CRAMP1
BTLA	NOL11	ZNF253	KREMEN1	CRYBG1
C2orf69	NUFIP2	ZNF257	LYPLA2	CSRNP3
C8orf58	NUP153	ZNF267	MAP2K6	CTNNBIP1
CAMKV	P2RY11	ZNF268	MCM2	CYP2C19
CCL8	PAFAH1B3	ZNF273	MDM2	DDI2
CCT5	PDCD6IP	ZNF275	MEST	DEK
CD302	PDIA6	ZNF281	MIXL1	DMXL1
CELF1	PEX26	ZNF319	MMP1	DNAJC28
CENPM	PIGM	ZNF426	MMP12	DUSP6
CEP57L1	PIK3R1	ZNF485	MMP14	EID2B
CNN2	PKM	ZNF550	MSH3	ERBB4
CNOT1	PLAG1	ZNF578	MTDH	FZD6
CNOT6	PNRC2	ZNF669	MTMR14	HBEGF
CNOT9	POM121C	ZNF682	MUC1	HIF1A
CRLS1	PPIC	ZNF701	MYC	HIST1H2AH
CSDE1	PSAP	ZNF704	MYO5A	HIST1H2BF
DCP2	PSMC3		MYO6	IGFBP5
DDX5	PTGFRN		MYOCD	IMMP2L

* Last accessed on 9 August 2025. Colored targets represent the strongly validated targets, meaning confirmed experimentally via Luciferase reporter assay, Western blot, and/or qRT-PCR.

). Focusing on the targets shared by at least two miRNAs (Figure 6), the STRINGapp Protein Query from Cytoscape 3.10.3 was employed to generate a physical subnetwork of protein–protein interaction (PPI). The top 20 enriched terms for Gene Ontology (GO) categories, DISEASES, TISSUES, KEGG pathways, and Reactome pathways were retrieved after conducting a functional enrichment analysis (Figure 7).

3. Discussion

Although the global incidence of CC has declined, it remains a major cause of cancer-related death in women, particularly when diagnosed at advanced stages, where treatment options are limited [2]. A personalized approach to CC management requires a comprehensive understanding of the tumor microenvironment, including how the disease interacts with the hemostatic system. VTE is a common complication in cancer, worsening prognosis, and in CC, it is associated with particularly poor survival [43,44,45]. This underscores the need for CC biomarkers to guide clinical decisions, while aiding VTE risk prediction [46]. In this regard, miRNAs recently emerged as promising candidates. Hence, this study investigated circulating miRNAs in CC patients undergoing concurrent chemoradiotherapy (CCRT), aiming to identify ncRNAs associated with both CAT and tumor aggressiveness to support personalized treatment strategies.

In this study, 13% of CC patients developed clinically evident VTE, though the absence of active screening suggests that the true incidence may be higher. This aligns with previous reports showing thromboembolic rates of 9-12% in CC patients, confirming their elevated thrombotic tendency [15,16]. Importantly, VTE was associated with significantly poorer survival, with affected patients showing a shorter mean OS than those without thrombotic complications (30.2 ± 4.3 months and 49.8 ± 2.0 months, respectively; log-rank test, p = 0.005), Consistently, VTE patients exhibited nearly a five-fold higher risk of death within 10 years (95% CI, 1.42–16.05; p = 0.011).

Among the miRNAs analyzed—miR-20a-5p, -23a-3p, -125b-5p, -145-5p and -616-3p—only miR-23a-3p and miR-145-5p showed a strong positive correlation (Pearson’s r = 0.700; p < 0.001), likely reflecting overlapping regulatory roles. Consistently, the in silico analysis further revealed that these miRNAs share multiple predicted target genes, which reinforces the likelihood of coordinated regulation in vivo. Consistently, both miRNAs have been previously implicated in the regulation of hemostasis and CC progression. In the general population with VTE, downregulation of miR-145-5p has been found to increase F3 expression, thus increasing TF levels, which promotes thrombus formation [47,48]. In turn, miR-23a is reported to target TFPI2, which encodes a regulatory protein involved not only in thrombin formation (inhibiting TF coagulation pathway) but also in platelet activation and fibrinolysis [49,50]. Parallelly, TFPI2 has been associated with CAT, where its high levels impair fibrinolysis, promoting the maintenance and progression of established thrombus. This was observed in epithelial ovarian cancer patients with asymptomatic VTE and a positive D-dimer result [31]. In CC, TFPI2 acts as a modulator of tumor progression. Namely, its downregulation through miR-23a is reported to contribute to angiogenesis in CC cells [51,52,53,54]. Similarly, miR-145-5p functions as a tumor suppressor by regulating oncogenic pathways [55]. Overall, the positive correlation observed between miR-23a-3p and miR-145-5p (Pearson’s r = 0.700; p < 0.001) likely reflects coordinated regulation of hemostasis within the tumor microenvironment, as they modulate the expression of complementary components of the coagulation pathway (TFPI2 and TF, respectively), with potential implications for both VTE risk and tumor progression.

Several moderate correlations were observed among the miRNAs, particularly between miR-20a-5p and miR-23a-3p (Pearson’s r = 0.659; p < 0.001), miR-20a-5p and miR-145-5p (Pearson’s r = 0.662; p < 0.001), and miR-125b-5p and miR-145-5p (Pearson’s r = 0.660; p < 0.001). These correlations are biologically plausible, since miR-20a-5p, miR-23a-3p and miR-145-5p all participate in the regulation of the TF pathway. Namely, miR-145-5p directly targets F3 [48], miR-23a-3p inhibits TFPI2 [49,56] and miR-20a-5p has been validated as another F3 regulator [57,58], with a previous in silico analysis suggesting that it may additionally target TFPI2 [56]. The shared roles of these miRNAs in TF-related processes likely underline their correlated expression. Regarding the correlation between miR-125b-5p and miR-145-5p, it may reflect their contribution to a pro-thrombotic microenvironment. Notably, miR-125b influences the expression of coagulation factor IX (encoded by the F9 gene) [59,60]. Both miR-125b-5p and miR-145-5p show decreased expression during HPV-driven cervical lesion progression, suggesting common regulation by viral infection [61,62,63]. However, their involvement in overlapping molecular pathways relevant to cervical carcinogenesis is poorly explored. As for miR-616-3p, it displayed no significant correlations, except with miR-20a-5p, possibly due to their shared TFPI2 modulation [47,56].

Considering both the entire cohort and subgroup analysis, miR-20a-5p and miR-145-5p emerged as the most relevant miRNAs in this study. Regarding miR-20a-5p, its low expression levels were significantly associated with VTE development in CC patients (OR = 0.12; 95% CI, 0.02–0.71), suggesting a role in CAT, consistent with its known function in suppressing TF levels [57,58]. Beyond coagulation, miR-20a-5p, a member of the oncogenic miR-17-92 cluster, has been linked to invasion and metastasis [64]. Previous reports indicate its higher expression in CC compared to healthy controls [65]. In this study, while no direct impact of miR-20a-5p expression pattern on CC patients’ OS was observed, the miRNA expression significantly varied depending on important clinicopathological features. Namely, reduced miR-20a-5p levels were more common in older patients (≥52 years) and those with CSCC tumors (ꭓ² test, p = 0.043 and p = 0.003, respectively). The first observation suggests an age-related difference in the miRNA regulation, highlighting age at CC diagnosis as a potentially relevant factor influencing miR-20a-5p expression. This is consistent with previous studies showing that the miR-17-92 cluster tends to be downregulated with aging, as its lower levels correlate with senescence-associated pathways [66]. Regarding the link between miR-20a-5p and CSCC tumors, though an underlying biological mechanism cannot be entirely dismissed, this finding was most likely influenced by the predominance of CSCC cases among patients who developed VTE. Notably, although no association between histological subtype and VTE was observed, all VTE cases occurred among CSCC patients.

MiR-145-5p expression pattern significantly contrasted depending on several clinicopathological features and survival outcomes in CC, but its role appears context- and histology-dependent. In this study, downregulation of miR-145-5p was more frequent among patients with CSCC tumors (ꭓ² test, p = 0.027). In agreement, a prior study showed reduced expression in both tissue and plasma of CSCC patients [63]. Interestingly, the same report linked lower miR-145-5p levels to advanced FIGO stages, while in this cohort, the miRNA downregulation was more common in patients with early-stage disease (ꭓ² test, p = 0.011). This discrepancy may be explained by differences in cohort characteristics. Further studies with larger cohorts are needed to clarify this finding.

Regarding patient survival, miR-145-5p expression showed no significant association with OS considering the entire cohort. However, stratified analysis focusing on ADC patients revealed that lower miRNA levels were associated with better outcomes (log-rank test, p = 0.039), indicating potential histology-specific effects. Similar patterns have been reported in esophageal cancer (EC), where miR-145 suppresses tumorigenesis in esophageal squamous carcinoma (ESCC) but promotes invasion and anoikis resistance in esophageal adenocarcinoma (EAC) [67]. This dual role may be explained by its regulation of c-Myc, a repressor of integrins. By targeting c-Myc, miR-145 indirectly increases integrin expression, which enhances tumor cell adhesion and invasion, ultimately contributing to a poorer prognosis [68]. Given that HPV infection is also a possible cause of EC, similar underlying mechanisms may also be relevant in CC. Additional investigation is needed to explore the histology-dependent expression of miR-145-5p in cervical tumorigenesis. In the VTE subgroup, patients with lower-to-medium miR-145-5p levels exhibited better OS compared to those with high levels (log-rank test, p = 0.018). Once again, this seemingly paradoxical finding may reflect histological subtype and disease stage distribution, as patients with lower miR-145-5p were more likely to have CSCC and earlier-stage tumors, both generally associated with better prognosis [7]. However, the prognostic impact of histology remains controversial. While several studies suggest that CSCC is associated with better prognosis compared to ADC, others report equivalent survival outcomes between these subtypes [69,70,71]. Overall, while miR-145-5p downregulation may not have directly influenced VTE development in our cohort, it appears to act as a surrogate marker of tumor characteristics and biological behavior, emphasizing its complex and context-dependent role in CC.

Concerning miR-23a-3p and miR-125b-5p, their expression levels were significantly associated with patient characteristics. Low-to-medium miR-23a-3p expression was more frequently observed in patients with early-stage CC (ꭓ² test, p = 0.025). While the relationship between circulating miR-23a-3p and CC stage is not fully established, prior studies suggest that miR-23a promotes CC progression by suppressing TFPI2, a factor whose inhibition is associated with increased tumor aggressiveness and more advanced disease of CC. Furthermore, high miR-23a levels have also been reported in metastatic fibroblasts [52]. Likewise, elevated miR-23a expression in gastric cancer correlates with advanced stage, lymph node metastasis and higher invasion [72], supporting its pro-tumorigenic role.

For miR-125b-5p, lower expression levels were more common in patients with CSCC (ꭓ² test, p = 0.018). This aligns with previous findings showing miR-125b-5p downregulation in HPV16-positive CSCC and high-grade squamous intraepithelial lesions compared to normal cervix tissue [73]. Similarly, in ESCC, reduced miR-125b-5p levels have been linked to HPV infection, and transfection with HPV16E6 is reported to decrease miR-125b-5p expression in ESCC cells [74]. Further investigation is needed to clarify the biological significance of these parallels between CSCC and ESCC.

Lastly, low plasmatic levels of miR-616-3p, along with low-to-medium levels of miR-23a-3p, showed a positive effect on patient OS in the ADC subgroup (log-rank test, p = 0.039 for both). This may be attributed to the specific biology of ADC, which typically exhibits higher vascular endothelial growth factor (VEGF) expression compared to CSCC, making VEGF a more relevant factor in this subtype [75]. VEGF promotes endothelial cell proliferation and angiogenesis but also induces TFPI2 expression as a negative feedback mechanism. TFPI2 inhibits endothelial cell migration and, at higher concentrations, can suppress VEGFR-2 activity and mitogen-activated protein kinase (MEK)/extracellular signal-regulated kinase (ERK) signaling, exerting anti-angiogenic effects [51,76]. VEGF-A overexpression in ADC has been linked to tumor aggressiveness and poorer prognosis [77]. In this study, the downregulation of miR-23a-3p and miR-616-3p in ADC patients was correlated with a less aggressive tumor phenotype. A possible explanation is that, since both miRNAs target TFPI2, their reduced levels could lead to an increase in TFPI2, expression which could counteract VEGF-driven angiogenesis. Consequently, higher TFPI2 activity could suppress angiogenesis, and, in this context, improve patient survival.

Within ADC cases, a co-expression pattern was observed for miR-23a-3p, miR-145-5p and miR-616-3p, with patients showing consistent expression levels across all three miRNAs. Given their similar associations with patient OS, their shared biological roles were explored in an in silico analysis. Pairwise comparisons of predicted targets showed that miR-23a-3p and miR-145-5p shared the most targets (N = 10), whereas miR-23a-3p and miR-616-3p, as well as miR-145-5p and miR-616-3p, each shared three targets. Functional enrichment analyses revealed that these shared targets were significantly associated with molecular functions, cellular components and biological processes relevant to tumor biology. DISEASE and TISSUE enrichment analyses further linked these targets to cervical carcinoma, the female reproductive system and vascular diseases. Additionally, Reactome and KEGG pathway analyses highlighted their involvement in oncogenic and pro-thrombotic signaling pathways, including transcriptional dysregulation in cancer, viral carcinogenesis and microRNAs in cancer. Collectively, these results suggest that the shared targets of miR-23a-3p, miR-145-5p and miR-616-3p converge on regulatory roles that integrate cervical tumorigenesis and potentially CAT. However, to confirm and complement the in silico analysis, in vitro and in vivo studies are mandatory, particularly considering the scarce evidence linking these miRNAs to CAT pathogenesis.

This study has some limitations that need to be acknowledged. First, as a retrospective study, it was constrained in its ability to collect relevant data, including comprehensive information on other potential risk factors such as oral contraceptive use, pregnancy and comorbidities. Additionally, VTE was not actively screened, raising the possibility of underdiagnosis, particularly for asymptomatic events. The absence of a healthy control group further limits the capacity to contextualize the observed associations against baseline physiological conditions. Also, this prevents the evaluation of whether the ncRNA dynamics studied are specific to CC-related CAT pathogenesis. Another limitation was the relatively small sample size, most likely making the study statistically underpowered, particularly when focusing on the number of VTE patients (N = 9). This is due to the rarity of both CC and CC-related VTE cases at our institution. These limitations emphasize the need for larger, prospective studies with standardized data collection and appropriate control groups to better elucidate the relationship between CC and venous thrombogenesis.

Despite the constraints, the study also presents interesting strengths. Namely, the cohort was relatively homogeneous, enhancing the internal validity of the findings and increasing confidence in the observed associations. Moreover, this study provides a novel contribution by focusing on circulating miRNAs in blood samples, rather than limiting analysis to CC tissues, as is common in most studies. By assessing miRNA profiles in circulation, this research highlights the potential of these highly stable molecules to be utilized in minimally invasive liquid biopsy approaches in CC patients, offering promising opportunities for disease risk assessment and prognostication.

4. Materials and Methods

4.1. Population Description

A retrospective cohort study was conducted involving adult patients from the North region of Portugal, all of whom had a confirmed diagnosis of CC. These individuals were admitted to the Clinic of Gynecology of the Portuguese Institute of Oncology of Porto (IPO Porto) between May 2017 and October 2021 to receive first-line treatment. Inclusion criteria comprised patients with CC classified between stages IB2 and IVA and those who were treated with CCRT as the primary treatment approach. The CCRT protocol consisted of weekly doses of cisplatin at 40 mg/m² administered during external radiotherapy with or without subsequent brachytherapy. Patients were excluded if they met any of these conditions: (1) younger than 18 years old, (2) had undergone surgery, considered a risk factor for CAT, at least six months prior to CCRT initiation, or (3) were only seeking a second medical opinion. In total, 69 patients were involved in this study, with a mean follow-up duration of 47.18 ± 2.11 months. All participants provided written informed consent aligned with the principles of the Helsinki Declaration. This study was approved by the ethics committee at IPO Porto (CES IPO:287A/014).

Demographic data, clinicopathological characteristics, and follow-up information were obtained from the patient’s medical records. Tumor staging was determined using the FIGO system [78]. CAT status was also established based on medical records. Notably, no active screening was conducted. A description of the study cohort characteristics is given in

Table 3. Demographic and clinicopathological characterization of CC patients (N = 69).

Variable	N (%)
Age at CC diagnosis (years)	51.9 ± 13.4 *
<52	33 (47.8)
≥52	34 (49.3)
Missing data	2 (2.9)
Histology
Squamous Cell Carcinoma	52 (75.4)
Adenocarcinoma	12 (17.4)
Other	3 (4.3)
Missing data	2 (2.9)
FIGO Stage
<IIB	9 (13.0)
≥IIB	58 (84.1)
Missing data	2 (2.9)
VTE Status
No	58 (84.1)
Yes	9 (13.0)
Missing data	2 (2.9)

* Mean ± standard deviation. Abbreviations: CC, Cervical Cancer; FIGO, International Federation of Gynecology and Obstetrics; N, number; VTE, Venous Thromboembolism.

.

4.2. Sample Collection and Processing

Before the initiation of CCRT protocol, peripheral blood samples of each patient were collected using a standard venipuncture method. Samples were obtained in ethylenediaminetetraacetic acid (EDTA)-coated vacuum tubes and centrifuged at 3000 rpm for 5 min at room temperature to isolate the plasma fraction. The plasma samples were then stored at −80 °C until the moment of RNA extraction.

4.3. MiRNA Selection

MiRNAs were selected based on an extensive literature review aimed at identifying the most relevant candidates implicated in VTE, both in the general population and among cancer patients [29]. In parallel, an exhaustive review of the existing literature on miRNAs involved in cervical carcinogenesis was conducted. MiRNAs were included based on the following criteria: (1) regulatory role in hemostasis; (2) documented involvement in VTE; (3) detection in blood circulation; and (4) established association with cervical carcinogenesis. Based on these inclusion parameters, hsa-miR-20a-5p, hsa-miR-23a-3p, hsa-miR-125b-5p, hsa-miR-145-5p, and hsa-miR-616-3p were selected.

MiR-20a-5p has been shown to regulate both F3 (gene encoding TF) [57] and TFPI2, thereby modulating hemostasis and thrombosis [56]. In CC, it is frequently upregulated in both tissue and serum samples [79,80], where it promotes tumor growth, invasion and autophagy, through the targeting of Runt-related transcription factor 1 (RUNX1) [81], Autophagy-related protein 7 (ATG7), Tissue Inhibitor of Metalloproteinase 2 (TIMP2) [65], Thrombospondin 2 (THBS2) [82], and Programmed cell death 6 (PDCD6) [83]. Similarly, miR-23a-3p and miR-616-3p also target TFPI2, influencing fibrinolysis and thrombus stability [31,49,50,56,84]. The former generally acts as an oncogene in several cancers, particularly in pancreatic carcinoma [49]. However, it is downregulated in CC, resulting in TFPI2 overexpression [85]. As for miR-616-3p, it has been linked to increased CAT susceptibility in ovarian cancer patients [56]. Yet, its role in CC has not been reported.

MiR-125b-5p has a recognized role in thrombotic disease, where its elevated circulating levels positively correlate with platelet count and ischemic stroke severity suggesting a pro-thrombotic function [86,87]; however, in cancer, miR-125b exhibits context-dependent roles, acting as both an oncogene and a tumor suppressor [88]. In CC, this miRNA is downregulated by HPV infection, a mechanism that facilitates viral persistence and carcinogenesis [61,89,90].

Another key regulator of the F3 is miR-145 [48]. In the general population (cancer-free individuals), reduced levels of this miRNA have been observed in VTE patients, likely reflecting activation of the coagulation cascade and thrombus formation [48,91]. Beyond TF, miR-145 has also been reported to target other hemostasis-related genes, including F11 (gene encoding FXI) [92,93] and SERPINE1 (gene encoding PAI-1) [94], where elevated plasma levels of these proteins’ expression have been associated with a higher risk of VTE occurrence [92,94,95]. In CC, miR-145 downregulation in both tissues and plasma correlates with poor prognosis and aggressive clinical features [55]. Functionally, miR-145 suppresses cervical carcinogenesis through targeting SMAD-Interacting Protein 1 (SIP1) [96], Helicase-Like Transcription Factor (HLTF) [97], Wnt Family Member 2B (WNT2B) [98], Regulators of Calcineurin 3 (RCAN3) [99] and Fascin actin-bundling protein 1 (FSCN1) [100]. Taken together, these five miRNAs converge on the regulation of coagulation and tumor progression, positioning them as strong candidates for the purpose of the study.

4.4. Total RNA Extraction

Before RNA extraction, previously stored plasma samples were thawed on ice and centrifuged for 4 min at 2000 rpm to remove residual cellular debris and platelets, as previously described [56]. Moving forward, total RNA was extracted from the samples using the MagMAX™ mirVana™ Total RNA Isolation Kit (CAT A27828, Thermo Fisher Scientific, Waltham, MA, USA), according to the manufacturer’s instructions. The extraction procedure was carried out using a KingFisher™ Duo Prime Magnetic Particle Processor (Thermo Fisher Scientific, Waltham, MA, USA). The concentration and purity of the extracted plasma RNA were assessed using a NanoDrop Lite spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). The RNA samples presented variable concentrations (8–25 ng/µL). Before cDNA synthesis, the samples were adjusted to a final concentration of 8 ng/µL. Regarding the RNA purity (A260/280 ratio), it varied between 1.20 and 1.35, which is expected when working with human plasma. After quantification, RNA samples were kept at −80 °C.

4.5. cDNA Synthesis

Complementary DNA (cDNA) was synthesized from 40 ng of RNA per sample using the TaqMan™ MicroRNA Reverse Transcription Kit (CAT 4366596, Thermo Fisher Scientific, Carlsbad, CA, USA), following the manufacturer’s instructions. Sequence-specific stem-loop reverse transcription primers were used for the following targets: hsa-miR-20a-5p (assay ID 000580), hsa-miR-23a-3p (assay ID 000399), hsa-miR-125b-5p (assay ID 000449), hsa-miR-145-5p (assay ID 002278), hsa-miR-451a (assay ID 001141), hsa-miR-616-3p (assay ID 002414), hsa-miR-1228-3p (assay ID 002919) and U6 snRNA (assay ID 001973). Each 10 μL reaction contained 5 μL of RNA and 5 μL of reaction mix. The latter included 10×RT Buffer, dNTP Mix, RT, RNAse inhibitor and the stem-loop RT primers. Reverse transcription was performed in Mycycler™ Thermal Cycler (Bio-Rad Laboratories, Hercules, CA, USA) under the following cycle conditions: 30 min at 16 °C, 60 min at 42 °C and 10 min at 85 °C [56]. Negative controls (lacking RNA) were included in all reverse transcription reactions to assess potential contamination.

4.6. Relative Quantification of miRNAs

Quantification of miRNA expression was performed by real-time quantitative polymerase chain reaction (RT-qPCR) using a StepOnePlus™ qPCR system (Applied Biosystems^®, Foster City, CA, USA). Each PCR reaction was carried out in a final volume of 10 μL, comprising 5.0 μL of 2× TaqMan™ Fast Advanced Master Mix (Applied Biosystems^®, Foster City, CA, USA), 2.5 μL of nuclease-free water, 0.5 μL of 20× specific TaqMan™ MicroRNA Assays for hsa-miR-20a-5p (assay ID 000580), hsa-miR-23a-3p (assay ID 000399), hsa-miR-125b-5p (assay ID 000449), hsa-miR-145-5p (assay ID 002278), hsa-miR-451a (assay ID 001141), hsa-miR-616-3p (assay ID 002414), hsa-miR-1228-3p (assay ID 002919) and U6 snRNA (assay ID 001973), along with 2.0 μL of cDNA template.

According to previous studies measuring miRNA expressions in human plasma samples, U6 snRNA and miR-1228-3p were tested as housekeeping genes [56,101]. Beyond its potential link to CAT and CC prognosis, miR-23a-3p—known for its stability in plasma samples upon hemolysis—and miR-451a—abundantly expressed in red blood cells—were employed as indicators to evaluate the presence of hemolysis [102]. A ΔCq value (hsa-miR-23a-3p–hsa-miR-451a) greater than 8 indicates a high likelihood of hemolysis [56,102,103]. Since three samples exceeded this threshold, only 66 were included in the miRNA expression analyses.

PCR amplification was performed under the following thermal cycling conditions: 2 min at 50 °C, 10 min at 95 °C, followed by 45 cycles of 15 s at 95 °C and 1 min at 60 °C. Target and reference gene expression levels were quantified for each sample on the same reaction plate. All quantifications were performed in triplicate, and negative controls (lacking cDNA) were included in every reaction for quality assurance. Cycle threshold (Ct) values with a standard deviation (SD) higher than 0.5 were not included in the analysis. The Thermo Fisher Connect platform (Thermo Fisher Scientific, Waltham, MA, USA) was used to set consistent baseline and threshold values across all plates.

4.7. Statistical Analysis

Data were analyzed using IBM SPSS Statistics for Windows (version 29, IBM Corp., Armonk, NY, USA), while graphical representations were generated with GraphPad Prism (version 9.0.0, GraphPad Software Inc., La Jolla, CA, USA). The Livak method was employed to determine the miRNA’s normalized relative expression. The miR-1228-3p was deemed the most suitable endogenous control, given its stable expression, which is consistent with a previous study of our research group using plasma samples [56].

Given the lack of previous studies exploring the role of the evaluated miRNAs in the context of CC-related CAT, the expression of each miRNA was analyzed considering four categorized expression profiles: (1) samples were divided into low and high expression groups using the median expression value as a threshold; (2) samples separated into three categories—low, intermediate, and high expression—based on terciles; (3) the first and second terciles were grouped as low expression and the third tercile was classified as high expression; and (4) the first tercile was deemed the low expression group, while the second and third terciles were combined as high expression. All these profiles were considered in analyses where miRNA expression was treated as a categorical variable. The same grouping method was previously applied in other studies of our research group [56,104].

Data normality was assessed using either the Shapiro–Wilk test or the Kolmogorov–Smirnov test, depending on the sample size (N ≤ 50 and N > 50, respectively). Based on the outcome of the normality test, correlations between miRNA expression levels were examined using either Pearson’s correlation coefficient (for normally distributed data) or Spearman’s rank correlation coefficient (for non-parametric data). Correlations were deemed noteworthy when the p-value was less than 0.05 and the correlation coefficient was ≥0.500. Correlation strength was interpreted using the following categories: coefficients between 0.500 and 0.700 were considered moderate, 0.700–0.900 strong, and ≥0.900 very strong [40,41,42].

The relationship between miRNA expression and patients’ demographic or clinical-pathological features (categorical variables) was evaluated by applying the chi-square test (χ² test). Age categories (<52 vs. ≥52 years) were based on the mean age upon CC diagnosis, given that the variable followed a normal distribution (Kolmogorov–Smirnov test, p > 0.05). Differences in miRNA expression according to VTE occurrence [no vs. yes] were assessed using either the Mann–Whitney U test or an unpaired t-test, depending on data normality. These comparisons were confirmed using the χ² test and logistic regression analysis.

The impact of miRNA expression on OS was assessed in the entire cohort and considering patients’ age (<52 vs. ≥52 years), histological type (CSCC vs. ADC vs. others), FIGO stage (<IIB vs. ≥IIB) and VTE status (no vs. yes). OS was defined as the period between CC diagnosis and patient death due to all causes or the last follow-up date. Kaplan–Meier survival analysis was used to generate survival curves, while differences between groups were evaluated using the log-rank test or Tarone-Ware test, depending on the proportionality and timing of events across groups. All tests were two-tailed with a 5% significance threshold.

4.8. In Silico Analysis

The biological relevance of the relevant miRNAs was investigated through a comprehensive in silico analysis. All validated miRNA targets were obtained using the miRTargetLink 2.0 database (https://ccb-compute.cs.uni-saarland.de/mirtargetlink2/; last accessed on 9 August 2025). PPI networks were then constructed via the STRINGapp Protein Query from Cytoscape 3.10.3, focusing on physical interactions. Functional enrichment analysis was subsequently performed, considering a false discovery rate (FDR) threshold of <0.01 and filtering out redundant terms with a similarity cut-off of 0.5. The top 20 enriched terms, if available, were retrieved for each category, including GO terms, disease (pathological processes), tissues (tissue-specific expression), KEGG pathways, and Reactome pathways.

5. Conclusions

CC patients are commonly affected by CAT, fueling tumor progression and worsening prognosis. Thus, identifying CC biomarkers to aid in decision-making is crucial [105,106]. MiRNAs are promising candidates due to their stability in liquid biopsies and their established roles in the regulation of both thrombosis-related pathways and CC biology [107,108]. MiRNA profiling could thus offer new opportunities for disease monitoring and therapeutic strategies in CC patients at increased thrombotic risk. Hence, this study evaluated the plasma expression levels of miRNAs involved in venous thrombogenesis and cervical carcinogenesis in a cohort of CC patients. According to the findings, VTE occurrence negatively affects the prognosis of CC patients. Among the miRNAs analyzed, miR-20a-5p and miR-145-5p demonstrated the most evident links with the studied diseases and clinical features. The former was the only miRNA significantly associated with CAT occurrence, suggesting its potential role as a predictive biomarker of thrombotic risk in CC. This could be useful to aid in thromboprophylaxis. Interestingly, besides its role in CAT, miR-20a-5p upregulation was linked to ADC histology, indicating a context-dependent function in cervical tumorigenesis. Regarding miR-145-5p, it exhibited a prognostic value in ADC patients and those with CAT events, also reflecting its context-specific role. Finally, in ADC patients, miR-23a-3p, miR-145-5p and miR-616-3p were found to have a similar influence on OS, highlighting their potential utility as prognostic biomarkers within this histological context. Overall, these preliminary findings emphasize the relevance of miRNA profiling as tools to better understand CC pathogenesis, assess thrombotic risk and improve prognostication. Moving forward, to complement the in silico analysis, future studies should aim to investigate the molecular mechanisms in CC and CAT involving these miRNAs through in vitro and in vivo studies, allowing the exploration of potential therapeutic targets. In conclusion, the evaluated miRNAs demonstrate functional significance and a predictive biomarker potential in CC, which could enable a more personalized risk assessment and targeted interventions.