The Combination of circEPSTI1 and MIF Offers Diagnostic Value for Endometrial Cancer

Introduction

Endometrial cancer (EC) is the third most prevalent malignancy among females worldwide,¹ yet no reliable approaches to detecting EC in its early stage currently exist for women at average risk who are free of any symptoms.² Under most established guidelines, initial detection strategies most commonly include endometrial biopsy or transvaginal ultrasonography, but both of these approaches are subject to limitations. Diagnostic curettage is also commonly employed when seeking to definitively identify this tumor type, but this procedure is inherently invasive and requires the scraping of tissue specimens without clear visibility such that missed diagnoses may occur, while also exposing patients to the potential for endometrial bleeding of varying severity levels. EC tumors can also invade the muscle layer, contributing to metastatic progression and adverse patient outcomes.³ This procedure is also highly specialized such that it can only be conducted by experienced gynecologists. As such, validated biomarkers that can more reliably aid in the diagnosis of EC are urgently needed in order to improve the prognosis of this form of cancer.

Circular RNAs (circRNAs) are a unique subset of transcripts that are extremely stable and resistant to degradation owing to their structural characteristics, defined by a closed covalent loop of RNA.^4–6 Particular circRNAs have increasingly been shown to regulate a wide array of cancers and other diseases through their ability to modulate biological processes such as differentiation, glycolytic metabolism, survival, proliferation, and migration.^7,8 These findings have spurred growing interest in the application of circRNAs as therapeutic targets or diagnostic biomarkers in a range of pathological contexts.^9,10 Widespread circRNA dysregulation has recently been reported in EC, with certain circRNAs serving as important regulators of disease-related processes.^11,12 Of note, circEPSTI1 can reportedly enhance certain oncogenic activities in cervical cancer,¹³ oral squamous cell carcinoma,¹⁴ ovarian cancer,¹⁵ non-small cell lung cancer,¹⁶ and breast cancer.¹⁷ How circEPSTI1 functions in EC, however, remains poorly understood. One recent study also determined that macrophage migration inhibitory factor (MIF) levels are closely associated with EC patient prognostic outcomes.¹⁸ While the bulk of prior MIF research has centered on inflammatory diseases,¹⁹ an increasing number of studies suggest that it may function as a key factor that links inflammation to oncogenesis owing to the importance of inflammatory activity and tissue remodeling in the context of cancer development.²⁰

Here, circEPSTI1 and MIF levels were compared between samples of serum obtained from patients with EC and healthy subjects, while also assessing these expression levels across a range of tumor stages. To more fully assess the link between these biomarkers and the biology of EC, the associations between circEPSTI1 and MIF levels and the clinicopathological characteristics of EC patients were additionally examined. Receiver operating characteristic (ROC) curves were also constructed to test their diagnostic performance. Overall, these findings provided strong evidence for the importance of circEPSTI1 and MIF in the pathogenesis of EC, while also suggesting that they can be leveraged to help diagnose this form of cancer.

Materials and Methods

Sample Collection

This study received approval from the Ethics Committee of the Affiliated Hospital of Hebei University of Engineering, and was performed in accordance with the principles of the Helsinki Declaration. Serum samples were obtained from patients with EC and healthy individuals (n=186 each) between May 2021 and December 2022. The informed consent were gained from the patients and their families. Blood samples were collected in RNAse-free tubes for RNA isolation, which was performed with 30 minutes after collection. Patient clinicopathological and molecular characteristics were additionally recorded.

qPCR

RNA was isolated from 500 uL of serum per patient with an RNA Isolation Kit (Vazyme Biotech), after which cDNA was synthesized with a Prime Script RT reagent Kit (Takara). Then, qPCR reactions were conducted with SYBR Green (Takara) by 7500 Fast Real-Time PCR System, utilizing the GAPDH for the normalization of expression with the 2^−ΔΔCt method. The primers used in this study were as follows: circEPSTI1, Forward (F), AAGCTGAAGAAGCTGAACTC, Reverse (R), CCGGTTTATATTTGGTGCTATC. GAPDH, F, GGAGCGAGATCCCTCCAAAAT, R, GGCTGTTGTCATACTTCTCATGG. The qPCR conditions used in this study were as follows: Initial 1 step at 95°C for 30 sec, followed by 40 cycles at 95°C for 5 sec and at 60°C for 34 sec.

ELISAs

MIF levels in appropriate samples were analyzed with an ELISA kit (R&D Systems) based on provided directions, with all samples being analyzed in triplicate.

Statistical Analysis

GraphPad Prism 8.0 (GraphPad Software) was used to analyze all data, which were compared with Student’s t-tests or one-way ANOVAs. Survival analyses were conducted as per the Kaplan–Meier method. P < 0.05 served as the cut-off when defining statistical significance.

Results

Participant Characteristics

For this study, plasma samples from 186 EC patients and an equal number of normal control individuals were analyzed. The characteristics of these patients are summarized in Table 1, and no differences in age were detected between these two groups of study subjects (P > 0.05).

Table 1 Clinicopathological Characteristics of EC Patients and Controls

CircEPSTI1 Levels in Serum Samples are Correlated with the Clinicopathological Characteristics of Patients with EC

To confirm the identity of circEPSTI1, the genomic loci of circEPSTI1 and the Sanger sequence of junction site of circEPSTI1 were detected by Sanger sequencing (Figure 1A). EC patient serum samples were found to exhibit higher levels of circEPSTI1 when analyzed via qPCR (Figure 1B). The EC patients were allocated to two groups according to the median circEPST11 level (n=93/group). Under these conditions, high circEPSTI1 expression in the serum was significantly related to the presence of stages (p=0.031), histologic grade (p=0.029), and residual tumor (p=0.022), as shown in Table 2. No significant relationships were seen with age (p=0.652), histological type (p=0.352), diabetes (p=0.186), hypertension (p=0.854), or menopause status (p=0.309), but patients expressing high circEPSTI1 did exhibit a reduction in both progression-free survival (PFS, Figure 2A) and overall survival (OS, Figure 2B) when compared to those patients exhibiting low expression levels. Consistently, univariate analyses detected a close relationship between the levels of this circRNA, stage, histologic grade and residual tumor, as shown in Tables 3 and 4. These associations remained significant when performing multivariate analyses.

Table 2 Relationships Between circEPSTI1 Expression Levels and Patient Clinicopathological Characteristics

Table 3 Univariate and Multivariate Cox Analyses of PFS-Related Factors in EC Patients

Table 4 Univariate and Multivariate Cox Analyses of OS-Related Factors in EC Patients

Figure 1 Patients with EC exhibit elevated serum levels of circEPSTI1. (A) The genomic loci of circEPSTI1 and the Sanger sequence of junction site of circEPSTI1. (B) Serum samples from EC patients and healthy subjects were assessed to compare serum circEPSTI1 expression, which was found to be elevated in individuals diagnosed with EC. ***P<0.001.

Figure 2 Serum levels of circEPSTI1 are related to the clinicopathological characteristics of patients with EC. Log rank tests were utilized to assess the PFS (A) and OS (B) of patients exhibiting low or high circEPSTI1 expression.

Serum MIF Levels Over Values for the Diagnosis of EC

There is prior evidence for an association between MIF concentrations and the prognosis of individuals with EC.¹⁸ Here, the mean respective concentrations of MIF in the serum of EC patients and healthy subjects were 4.825±1.05 ng/mL and 5.871±3.37 ng/mL (P < 0.0001) (Figure 3A). As revealed by constructed ROC curves, levels of serum MIF were capable of reliably distinguishing between patients with EC and healthy individuals, with an Area of 0.6646 (95% CI; 0.6086–0.7205) (Figure 3B).

Figure 3 Evaluation of the expression and diagnostic value of MIF in patients with EC. (A) Serum MIF concentrations were elevated in EC patients as compared to healthy subjects. (B) Serum MIF concentrations were able to distinguish between EC patients and control subjects in ROC curve analyses. ***P<0.001.

EC Can Be Effectively Diagnosed Based on Serum circEPSTI1

To similarly assess the diagnostic performance of the serum levels of circEPSTI1, ROC curves were again constructed, effectively allowing for differentiation between samples from individuals with EC and healthy subjects. At the optimal cut-off of 3.9496, the Area for the ROC curve was 0.8351 (95% CI: 0.7941–0.8761) (Figure 4A). CircEPSTI1 may thus outperform MIF with respect to its sensitivity when leveraged as a biomarker to diagnose EC. These analyses were further expanded by examining the diagnostic performance of a combination of serum circEPSTI1 and MIF as a means of detecting EC. The combined AUC when using these two biomarkers was 0.8604 (95% CI: 0.8217–0.8991) (Figure 4B). Predictive probability levels above this cut-off were indicative of EC positivity.

Figure 4 Serum circEPSTI1 offers diagnostic value for the detection of EC. (A) Serum levels of circEPSTI1 were able to effectively distinguish between EC patients and controls when constructing an ROC curve. (B) An ROC curve analysis was conducted for the combination of circEPSTI1 and MIF to more effectively distinguish between healthy subjects and individuals with EC.

Discussion

EC is among the most frequently diagnosed malignancies among women. While many patients are diagnosed while the disease is in its early stages, the overall survival of individuals with recurrent disease or who are only diagnosed when the disease is relatively advanced is substantially shorter. Analyses of biomarkers in the peripheral blood thus hold great promise as a noninvasive and cost-effective means of aiding in EC diagnosis.

Many studies have emphasized the value of circRNAs in the diagnosis of a variety of cancers as these molecules are both highly stable and abundant, in addition to playing important roles in key cancer-associated processes.^21,22 The stability of circRNAs in patient serum and plasma remains consistent even with freezing and thawing or shifts in pH,^23,24 making them optimal minimally invasive targets for clinical assessment. Serum Circ-FAF1/Circ-ELP3 offers great utility as a biomarker when applied in the context of diagnosing breast cancer,²⁵ while hsa_circ_0000437 can promote gastric cancer pathogenesis and metastasis to the lymph nodes.²⁶ Levels of hsa_circ_0000702 in patient serum have been proposed to be an effective gastric cancer-related biomarker.²⁷ Although identified only recently, circEPSTI1 has been noted to support tumorigenesis and progression in cervical cancer,¹³ oral squamous cell carcinoma,¹⁴ ovarian cancer,¹⁵ non-small cell lung cancer,¹⁶ and breast cancer.¹⁷ The functions and diagnostic performance of circEPSTI1 in EC, however, have not been previously documented.

This study is the only one we are aware of examining the relevance of MIF in EC patients. MIF has also been tied to angiogenic activity and the induction of tumor progression and growth in various cancers, although not in EC. In patients with malignant ascites, for example, Hagemann et al noted strong MIF expression, while ovarian cancer cell-derived MIF was capable of stimulating cytokine, chemokine, and pro-angiogenic factor production, contributing to angiogenic activity and vascularization.²⁸ Nishihira et al further highlighted a close association between MIF and both angiogenesis and tumor growth through studies in which they used antisense MIF to treat murine colon cancer cells.²⁹ These results support the use of MIF as a biomarker for diagnosing EC.

Here, serum circEPSTI1 levels in EC patients were evaluated to gauge their diagnostic value. These analyses revealed that higher circulating serum levels of this circRNA were related to EC patient stages, histologic grade, and residual tumor. This suggests that circEPSTI1 may play a role in the development or malignant progression of this form of cancer. The OS and PFS of patients expressing higher circEPSTI1 levels were worse than those of patients in which this circRNA was not highly abundant, further underscoring its potential application in the context of patient management. The ROC curves showed that serum circEPSTI1 levels could reliably differentiate between patients with EC and healthy serum samples at high levels of sensitivity and specificity better than those for MIF alone. When combining serum circEPSTI1 and MIF analyses together, EC could be accurately diagnosed with even greater specificity and sensitivity, indicating that this combined biomarker panel may be a robust strategy for EC diagnosis in the clinic.

Conclusion

In conclusion, EC patients exhibit pronounced circEPSTI1 upregulation that may be closely tied to oncogenic activity such that efforts to measure circulating levels of this circRNA can aid in the more effective and timely diagnosis of this disease. The combination of circEPSTI1 with a range of other clinical biomarkers that are part of standard analyses may further enhance its diagnostic performance. While these results are subject to some limitations, they nonetheless underscore important opportunities for research focused on assessing the clinical utility of circEPSTI1 in combination with MIF when seeking to screen for EC.