Home Construction and validation of a diagnostic model for cholangiocarcinoma based on tumor-educated platelet RNA expression profiles
Article Open Access

Construction and validation of a diagnostic model for cholangiocarcinoma based on tumor-educated platelet RNA expression profiles

  • Haiyang Hu , Jiefeng He EMAIL logo and Haoliang Zhao EMAIL logo
Published/Copyright: January 13, 2025
Download Article (PDF)
Oncologie
From the journal Oncologie Volume 27 Issue 2

Abstract

Objectives

We aim to explore the diagnostic value of platelet-based “liquid biopsy” technology for cholangiocarcinoma (CCA), seeking reliable methods for early cancer diagnosis to improve patient prognosis.

Methods

Bioinformatics methods were utilized to analyze the GEO databases (GSE183635) and (GSE68086), identifying differentially expressed genes and constructing a diagnostic model of CCA using tumor-educated platelet (TEP) RNA expression profiles. GO and KEGG pathway enrichment analysis were performed. Additionally, platelet RNA from CCA patients and controls totaling 60 cases was extracted by qRT-PCR experiments to validate the diagnostic reliability of candidate genes, further confirmed through in vitro experiments.

Results

A diagnostic model comprising seven platelet genes (CRYM, IFI27, EED, METAP1, RASGRP1, SEC11A, and WDR82) effectively distinguished CCA from controls. Area under curve (AUC) values were 0.862 (training set), 0.875 (internal validation), 0.865 (total internal), and 0.954 (external validation). GO analysis highlighted “non-coding RNA processing,” “nuclear envelope,” and “catalytic activity, acting on RNA.” KEGG pathways included “Ribosome biogenesis in eukaryotes”, “RNA translocation” and “Spliceosome”. qRT-PCR experiments revealed significant differences (p<0.05) in METAP1, SEC11A, WDR82, RASGRP1, and EED gene expression in CCAs, consistent with bioinformatics predictions. CRYM showed significant differences (p<0.001) compared to healthy individuals. WDR82 and CRYM had high diagnostic efficiency (AUC 0.939 and 0.942), surpassing conventional tumor markers (AFP, CEA, and CA19-9). Joint receiver operating characteristics (ROC) analysis yielded an AUC of 0.806, sensitivity of 1.000, and accuracy of 0.833.

Conclusions

Based on the GEO database, we identified seven TEP RNAs (CRYM, IFI27, METAP1, SEC11A, WDR82, RASGRP1, EED) with strong discriminative ability for CCA, suggesting their potential as reliable non-invasive biomarkers.

Keywords: cholangiocarcinoma; liquid biopsy; tumor-educated platelet; diagnostic model RNA

Introduction

In recent years, there has been a progressive annual increase in the incidence and related mortality rates of cholangiocarcinoma (CCA), also known as bile duct cancer (BC) [1]. Recognizing the significance of early diagnosis and intervention becomes pivotal in enhancing patients’ prognoses and overall survival outcomes. Nevertheless, the conventional pathological biopsy, currently considered the gold standard for cancer diagnosis, has inherent limitations including invasiveness, limited reproducibility, and the ability to capture only a single temporal and spatial snapshot, which severely limit its utility in the early detection of tumors [2], [3], [4], [5]. The accuracy of current imaging tests and circulating tumor biomarkers for CCA diagnosis is far from satisfactory [6]. Furthermore, the high heterogeneity of CCAs at the genomic, epigenetic, and molecular levels severely compromises the efficacy of the available therapies [7]. These shortcomings have driven the field of oncology to shift its research focus toward liquid biopsies [8].

Blood-based “liquid biopsies”, which are minimally invasive, easily accessible, highly reproducible, free of potential complications, and highly effective in monitoring disease and tracing tumor resistance, have been widely demonstrated to be a viable alternative for noninvasive evaluation of tumor-specific biomarkers [9]. Serum extracellular vesicles (EVs) contain protein biomarkers for the prediction, early diagnosis, and prognostication of CCA that are detectable using total serum, representing a tumor cell-derived liquid biopsy tool for personalized medicine [6].

Tumor-educated platelet is a developing notion involving the transmission of tumor-related biomolecules to platelets, leading to their being educated. Platelets are small pieces of cytoplasm shed from the cytoplasmic lysis of mature megakaryocytes in the bone marrow and are a formative component of the blood. Besides their well-known functions – thrombosis and hemostatic regulation – platelets are also an important component of the immune system, participating in innate and adaptive immune responses, atherosclerosis, angiogenesis, and lymphatic vessel development [10]. Increasing evidence suggests that interactions between tumor cells and circulating platelets are involved in regulating tumorigenesis, angiogenesis, dissemination, and metastasis. Platelets primarily influence cancer progression through several mechanisms [11], [12], [13]: firstly, they can aggregate around tumors, thereby bolstering tumor growth and aiding in evading immune surveillance; secondly, they facilitate the adhesion of tumor cells, aiding in their evasion of immune-mediated destruction; thirdly, the activation of platelets serves to promote tumor cell invasion and metastasis through processes involving the synthesis of lipid products, the release of proteins from α granules, and the induction of crucial events like epithelial-mesenchymal transition (EMT), vascularization, resistance, and exocytosis within tumor cells. On the other hand, cancer-induced platelets become activated and thus function [14], primarily through the direct – By transferring tumor-derived RNA or indirect- By releasing signals that regulate platelet mRNA processing induction of changes in platelet transcriptome profiles by tumor cells. In the effect of tumor cells and the tumor microenvironment, immature platelet mRNAs are stimulated to transform into mature mRNAs, which are subsequently translated into functional proteins, giving rise to tumorigenic platelets.

Although the clinical application of tumor-educated platelet (TEP) mRNA profiling has proven to be a crucial tool for cancer prediction, diagnosis, prognosis, and monitoring of biomarkers [15], the potential for use in CCA remains largely unexplored. In this study, we aim to construct a diagnostic model for CCA through mining and bioinformatics analysis of the GEO database. The objective was to explore whether TEP mRNA is meaningful for the diagnosis of CCA. Next, qRT-PCR experiments were conducted to validate the findings, providing a basis and ideas for the discovery of novel biomarkers for CCA.

Materials and methods

Database data and bioinformatics analysis

Data sources and identification of differential expression platelet genes

The platelet RNA expression sequencing datasets (GSE183635, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE183635) and GSE68086 https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi) of CCA patients and control groups were mined and downloaded in the GEO database, including 85 cases of CCA and 390 cases of the normal control group in GSE183635, and 14 cases of CCA and 55 cases of normal control group in GSE68086. All CCA and healthy control data were available and included in the study. The GSE183635 dataset was divided into a training set (70 %) and an internal validation set (30 %), while GSE68086 was used as an external validation set. The diagnostic model was trained and validated. Differentially expressed genes (DEGs) between CCA patients and healthy controls were identified using the “limma” R package (3.52.2) after data normalization and log base two conversion of gene read counts to transcripts per million normalized (TPM). DEGs are defined as false discovery rate (FDR) <0.05 and |log FC| >1. Volcano maps representing differentially expressed genes (DEGs) are plotted using the R packages “ggplot2” (3.4.4) [16].

Construction of a diagnostic model based on differentially expressed platelet genes (DEPGs) in cholangiocarcinoma (CCA)

In order to facilitate gene detection in TEP, differentially expressed platelet genes (DEPGs) were further screened according to three conditions: the average expression of DEPGs in CCA and normal groups ranked in the top 10 % of all DEPGs, and the FDR of DEPGs ranked in the top 100. Next, Lasso regression analysis was performed on genes that met these three conditions using the R package (“glmnet” (4.1.7)) to find the best regression coefficients and genetic scores. This process aims to construct a multigenic optimal diagnostic model for platelet RNA expression profiles for CCA. Finally, the R package “pROC” (1.18.0) is used to analyze the Receiver Operating Characteristic (ROC) curve of the model to quantitatively evaluate the diagnostic performance of the model.

Verification of diagnostic models

The model validation process involved using an internal validation set of the GSE183635 dataset (30 %), the complete data set from GSE183635 (100 %), as well as the GSE68086 dataset for external validation. To verify whether the diagnostic model has a significant value in the diagnosis of CCA, we plotted the ROC curve of the validation dataset and calculated the area under curve (AUC) value. P<0.05 was considered statistically significant.

GO and KEGG pathway enrichment analysis

In order to further elucidate the potential biological functions of the differential genes screened out in platelets, the R package “clusterProfiler” (4.4.4) was used to conduct GO (Gene Ontology term) and KEGG pathway (Kyoto Encyclopedia of Genes and Genomes) enrichment analysis [17]. P<0.05 was considered a significant level of enrichment.

Protein-protein interaction (PPI) network construction and identification of hub gene

The PPI network was constructed from an online STRING database (https://string-db.org), and the overall score>0.7 was considered to be statistically significant. The molecular interaction network was visualized using Cytoscape 3.10.3, an open-source bioinformatics software platform. The hub genes were identified using the Degree algorithm from the cytoHubba plugin within Cytoscape. To explore the biological roles of these hub genes within CCA-associated modules, GO and KEGG analyses were also conducted using the R package “clusterProfiler” (4.4.4). A significance threshold of p<0.05 was considered for these analyses.

Collection of patients’ data and verification of candidate genes’ expression levels

Collection of clinical specimens

According to the standards of the World Health Organization, 18 blood samples from CCA patients, 27 blood samples from patients with non-cholangiocarcinoma biliary diseases (patients with bile duct stones) who were admitted to Shanxi Bethune Hospital’s Hepatobiliary Surgery Department and 15 blood samples from healthy volunteers in the physical examination center were collected from November 2022 to June 2023. Meanwhile, the clinical medical records of patients were collected. This research protocol was approved by the Medical and Health Ethics Review Committee of Shanxi Bethune Hospital (approval number: YXLL-2023-170). All participants read the study protocol and signed an informed consent form before being recruited to participate in the study. All participants must meet the inclusion and exclusion criteria.

Inclusion and exclusion criteria

Inclusion Criteria for the Experimental Group: firstly, patients preliminarily diagnosed with bile duct cancer through tumor markers and imaging findings, and subsequently confirmed with histological biopsy; secondly, patients who have not undergone any anti-tumor treatments such as chemotherapy, radiation therapy, immunotherapy, targeted therapy, or surgical intervention due to disease progression or objective reasons.

Exclusion Criteria: firstly, patients with concurrent other malignant tumors; secondly, patients diagnosed with platelet-related disorders such as Bernard-Soulier syndrome, thrombocytopenia, thrombocytosis, and other platelet abnormalities (all cases confirmed through postoperative histopathology or puncture biopsy). The control group selected patients with bile duct stones admitted to the Department of Hepatobiliary Surgery of Shanxi Bethune Hospital and healthy examiners in the physical examination center.

Isolation and extraction of platelets (modified gradient centrifugation)

All patients, either at the time of diagnosis or prior to surgery, as well as healthy volunteers, were provided a fasting peripheral blood sample of 7 mL using EDTA anticoagulant tubes. Subsequently, platelet RNA extraction was performed at room temperature. Ideally, blood samples were processed within 2 h. According to instructions provided by the manufacturer Solarbio, platelets were extracted from blood using the Human Peripheral Blood Platelet Isolate Kit (Lot No.: P6390, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China). The samples were centrifuged at 120 × g for 10 min to isolate platelet-rich plasma from the supernatant. Following this, a second centrifugation at 200 × g for 15 min was carried out to remove any remaining red blood cells and white blood cells from the plasma. Platelet counts were determined using an optical microscope (DMi8, Leica Microsystems, Wetzlar, Germany) to verify the successful isolation and extraction of platelets. Then, the platelets were washed twice at room temperature with a centrifugation at 500 × g for 20 min each, to remove the washing solution and obtain a platelet pellet.

Purity detection of platelets

Platelet RNA is much less than leukocytes, so the number of leukocytes should be reduced as much as possible to avoid contamination of platelet RNA profiles. Plasma smears were prepared using whole blood and subjected to a 2 h natural sedimentation and gradient centrifugation technique. These smears were then stained using the Wright’s staining solution (Lot No.: C0135-100 mL, Beyotime Biotechnology Co., Ltd, Shanghai, China). Next, platelets and leukocytes were quantified under high-magnification microscopy (DMi8, Leica Microsystems, Wetzlar, Germany). The platelet suspension purity achieved through gradient centrifugation, with 0–5 leukocytes per 106 platelets, was confirmed to meet the acceptable standards for quality platelet extraction [18].

Extraction of total RNA from platelets and real-time fluorescent quantitative reverse transcription polymerase chain reaction (qRT-PCR)

According to the instructions provided by the manufacturer Mei5 Biotech, the M5 Quickspin Whole Blood RNA Rapid Isolation kit (Lot No.: MF612-01, Mei5 Biotechnology Co., Ltd., Beijing, China) was used for total RNA extraction. Then, the extracted platelet RNA was evaluated for quality. After diluting the obtained RNA sample stock solution 50-fold, 1 μL was spotted on a μDrop ultra-micro detection plate (Catalog No.: N12391, Thermo Fisher Scientific, CA, USA), and the concentration and purity of RNA were detected with the spectrophotometer (Spectrophotometer 1,530, Thermo Fisher Scientific). Samples with suitable purity (D260 ratio between 1.8 and 2.2) were considered valid, and the RNA concentration was recorded for subsequent experiments. Subsequently, the total RNA mentioned above was reverse transcribed into cDNA using the M5 Super qPCR RT Kit with gDNA remover (Lot No.: MF166-plus-T, Mei5 Biotechnology Co., Ltd.). Among them, the reverse transcription mixed reaction solution is as follows: The reaction mixture consisted of 4 × DNA Remover Mix (including gDNA remover) 2 μL, RNA template 0.01–0.05 μg, and DEPC-treated water added to reach a final volume of 10 μL. The reaction mixture was gently mixed and incubated at 37 °C for 5 min. Subsequently, 5 × M5 RT Super Mix was added to the reaction mixture, and the reverse transcription reaction was performed using the BIO-RAD C1000 TOUCH PCR cycler (Bio-Rad Laboratories‌, CA, USA). The reaction conditions were set as follows: 42 °C for 15 min, followed by 50 °C for 5 min. At the end of the reaction, the synthesized cDNA samples were either kept on ice for immediate use or stored at −80 °C for future use, followed by Real-time PCR analysis. The PCR primers were obtained from SBS Genetech Co., Ltd (Shanghai, China) (Supplementary Table S1). The detection system utilized was the BIO-RAD CFX Connect Real-Time PCR system (manufactured by BIO-RAD). The 20 μL system was formulated as follows: 0.4 μL each for CRYM, IFI27, EED, METAP1, RASGRP1, SEC11A, WDR82, GAPDH forward and reverse primers (10 μM); The volume of cDNA products varies according to the number of copies of the target gene; 2 × M5 HiPer SYBR Premix EsTaq (with Tli RNaseH) 10 μL; Add DEPC water to a total volume of 20 μL. During the PCR amplification reaction, the double-stranded DNA was predenaturated for 30 min at 95 °C, denatured for 15 s at 95 °C, annealed and extended for 1 min at 60 °C, and the above was one cycle, with a total of 40 cycles of amplification.

Set a uniform Ct threshold (>35 is considered invalid) to analyze experimental results, and set three complex wells per group. The PCR automatically collects fluorescence signals during the reaction, records the Ct values of candidate genes and internal reference genes, and then uses relative quantification to determine the expression of candidate genes relative to internal reference genes. In this study, GAPDH was used as the mRNA normalized control, and there is sufficient evidence that this gene is stably expressed in platelets [19]. The relative expression level of platelet RNA was calculated by formula 2−ΔCt, where ΔCt=Ct (target gene) – Ct (internal reference gene).

In vitro experiments to verify changes in candidate genes

The human CCA cell line HUCC-T1 (Lot No.: STCC10101P, Wuhan Zishan Biotechnology Co., Ltd., Wuhan, China) was cultured in RPMI-1640 complete medium supplemented with 10 % fetal bovine serum (FBS) (Lot No. PM150110B, Procell Life Science & Technology Co., Ltd., Wuhan, China), while the human immortalized liver bile duct cell line THLE-3 (Lot No.: ZK0490(XR), Beijing Zhongke Quality Inspection Biotechnology Co., Ltd, Beijing, China) was cultured in Dulbecco’s modified Eagle’s medium (DMEM; Lot No.: PM150210, Procell Life Science & Technology Co., Ltd.) supplemented with 10 % FBS (Lot No.164210, Procell Life Science & Technology Co., Ltd.). Cell STR identification is correct and free of mycoplasma contamination. Both cell lines were maintained in a constant temperature incubator at 37 °C with 5 % CO2 under controlled conditions. The cell culture process strictly followed aseptic techniques to prevent contamination. When a monolayer of cells covering over 90 % of the culture dish is observed under a microscope, the cells will proceed to be collected for subsequent qRT-PCR experiments, following the same procedures as described in the section “Extraction of total RNA from platelets and real-time fluorescent quantitative Reverse Transcription polymerase chain reaction (qRT-PCR)”.

Sample storage, cell culture, RNA extraction, cDNA synthesis, and qRT-PCR analysis were all done at the Shanxi Provincial Key Laboratory of Hepatobiliary and Pancreas.

Statistical methods

Clinical data sheets were analyzed using the R “stats” package (4.2.1). If the variable is numeric, the One-way ANOVA test is used when the data satisfies the normal distribution and satisfies the homogeneity of variance test. When the data satisfies the normal distribution but does not satisfy the homogeneity of variance test, the Welch one-way ANOVA test is used; The normal distribution is not satisfied, and the Kruskal–Wallis test is used. If the variable is categorical and meets the conditions of theoretical frequencies greater than five and a total sample size of at least 40, inter-group comparisons are conducted using the chi-square test. When the data satisfies the conditions of 1–5 for theoretical frequencies and a total sample size of at least 40, inter-group comparisons are performed using the Yates’ correction. Inter-group count data were represented using frequencies (or frequencies), while measurement data were presented as mean ± standard deviation or median ((interquartile range (IQR)). The diagnostic efficacy of seven candidate genes was analyzed using ROC. A p-value of less than 0.05 was considered statistically significant for detecting differences.

Results

Identification of platelet difference genes

Download platelet RNA sequencing data for both CCA and healthy samples from the GEO database. The dataset GSE183635 yielded a total of 1,204 differentially expressed platelet genes, while GSE68086 yielded a total of 3,421 differential genes. After performing analyses using the “limma” R software package on the training set (70 %, containing 900 differential genes), internal validation set (30 %, containing 1,663 differential genes), combined dataset (100 %), and external validation set, a total of 393 differentially expressed platelet genes were identified (Figure 1E). The volcano map of each dataset are shown in Figure 1.

Figure 1: Initial screening of differentially expressed genes (A) Volcano map representation of training set (GSE183635 70 %) (B) internal validation set (GSE183635 30 %) (C) internal total set (GSE183635 100 %) (D) external validation set (GSE68086) (E) the number of differentially expressed genes produced by the intersection of each dataset.
Figure 1:

Initial screening of differentially expressed genes (A) Volcano map representation of training set (GSE183635 70 %) (B) internal validation set (GSE183635 30 %) (C) internal total set (GSE183635 100 %) (D) external validation set (GSE68086) (E) the number of differentially expressed genes produced by the intersection of each dataset.

Development and verification of diagnostic models

The “LASSO” R software package was employed to conduct regression analysis on the differential genes, identifying the optimal LASSO regression coefficients and determining the most significant combination of differentially expressed genes, thus successfully constructing a diagnostic model related to bile duct cancer (Figure 2A). Therefore, we obtained seven platelet genes, including CRYM, IFI27, EED, METAP1, RASGRP1, SEC11A, and WDR82 (Figure 2B). Seven platelet genes were identified according to the model, and the formula was: index=CRYM × (−0.03410378) + EED × (0.03245786) + IFI27 × (−0.01180892) + METAP1 × (0.03738255) + RASGRP1 × (0.01964733) + SEC11A × (0.04254028) + WDR82 × (0.02176244). The joint AUC for the most significant differential gene was 0.862 (Figure 2C) in the training set, 0.875 (Figure 2D) in the internal test set, and 0.865 (Figure 2E) shown in the total set. At the same time, the external dataset also showed good diagnostic ability, with an AUC of 0.954 (Figure 2F). As a result, it is believed that the diagnostic model has good diagnostic efficacy and can be used for the diagnosis of CCA.

Figure 2: Screening and testing of the most significantly different genes (A) Use LASSO regression analysis to find the minimum number of covariates corresponding to them (B) Changes in the expression of the most significant difference genes were obtained by screening in the normal group and the cholangiocarcinoma (CCA) cancer group (C–F) ROC joint curve for (C) training set (GSE183635 70 %) (D) internal validation set (GSE183635 30 %) (E) internal total set (GSE183635 100 %), and (F) external validation set (GSE68086). ***p<0.001.
Figure 2:

Screening and testing of the most significantly different genes (A) Use LASSO regression analysis to find the minimum number of covariates corresponding to them (B) Changes in the expression of the most significant difference genes were obtained by screening in the normal group and the cholangiocarcinoma (CCA) cancer group (C–F) ROC joint curve for (C) training set (GSE183635 70 %) (D) internal validation set (GSE183635 30 %) (E) internal total set (GSE183635 100 %), and (F) external validation set (GSE68086). ***p<0.001.

GO and KEGG pathway enrichment analysis

To explore the potential functions of candidate genes, GO and KEGG pathway enrichment analyses were performed. GO analysis of these differentially expressed platelet genes revealed that “non-coding RNA processing”, “nuclear envelope”, and “catalytic activity, acting on RNA” were the most common biological terms for biological processes, cellular components, and molecular functions, respectively (Figure 3A–C). KEGG analysis showed that these platelet genes were mainly enriched in “ribosomal biogenesis in eukaryotes”, “RNA transport”, “spliceosomes”, and “Th1 and Th2 cell differentiation” (Figure 3D). Obviously, the key genes screened are closely related to tumor proliferation activities, and the differentiation process of Th1 and Th2 cells is closely related to tumor immune response.

Figure 3: Function enrichment analysis of platelet-related genes (A) Top 10 significantly enriched biological processes of GO (B) Top 10 significantly enriched cell components of GO (C) Top 10 significantly enriched molecular functions of GO (D) Top 10 significantly enriched KEGG pathways. The size of the points represents the number of genes annotated to the pathway, and the color represents significance.
Figure 3:

Function enrichment analysis of platelet-related genes (A) Top 10 significantly enriched biological processes of GO (B) Top 10 significantly enriched cell components of GO (C) Top 10 significantly enriched molecular functions of GO (D) Top 10 significantly enriched KEGG pathways. The size of the points represents the number of genes annotated to the pathway, and the color represents significance.

Protein-protein interaction (PPI) network construction and hub gene identification

A PPI network of differential genes was constructed through the STRING database and visualized using Cytoscape (Figure 4A). The “Degree” algorithm of the CytoHubba plug-in obtained the PPI network and the first 20 hub genes identified from it (Figure 4B), and the correlation analysis was performed (Figure 4D). GO and KEGG analysis showed that the Hub gene was mainly concentrated in four biological processes: “ribosome biogenesis, rRNA metabolism process, rRNA processing, and ribosome biogenesis in eukaryotes” (Figure 4C). This enrichment result was consistent with the above description, which jointly verified the correlation between the screened differential genes and tumor proliferation.

Figure 4: Construction and functional enrichment analysis of PPI network (A) PPI network of differential genes (B) Screening of hub genes and protein interaction mapping (C) Functional enrichment analysis of hub genes (D) correlation analysis of hub genes. *p<0.05, **p<0.01.
Figure 4:

Construction and functional enrichment analysis of PPI network (A) PPI network of differential genes (B) Screening of hub genes and protein interaction mapping (C) Functional enrichment analysis of hub genes (D) correlation analysis of hub genes. *p<0.05, **p<0.01.

Sample quality control

To confirm the purity of isolated platelets, freshly isolated platelet samples were randomly selected for morphological analysis. Platelet-rich plasma (PRP) obtained by natural sedimentation plasma and gradient centrifugation were stained with Wright and observed under a microscope. As shown in Supplementary Figure S1, it was observed that the naturally settled plasma obtained had varying numbers of white blood cells at 200 × and 400 × high-power field views. In contrast, the platelet-rich plasma obtained through gradient centrifugation showed no white blood cells in both high-power field views and no red blood cells were observed. This is consistent with previous research findings, indicating that the proportion of nucleated cells in every 10 million platelets is less than 5, which can be considered as achieving high purity of platelets [20]. Next, the extracted platelet RNA was evaluated for quality. Samples with appropriate purity will have their RNA concentrations recorded for subsequent experiments.

Analysis of general baseline data of patients

To test the ability of candidate genes to discriminate in the diagnosis of CCA, we collected blood from a total of 60 patients, including 18 patients with CCA, 27 with bile duct stones, and 15 healthy people. The general clinical baseline features of the participants are shown in Table 1. There were no significant differences between the BC group and the BS and NC groups in terms of sex and age distribution, whether they smoked, and whether they had hepatitis virus infection (p>0.05). Participants in the BC group tended to drink alcohol before or present compared with the two control cohorts (p=0.024), a result consistent with our previous understanding that alcohol consumption is an independent risk factor for liver cancer and CCA [21]. Recent research has demonstrated that potential risk factors for BC include choledocholithiasis, cholelithiasis, HBV (Hepatitis B Virus) infection, advanced age, male gender, diabetes, smoking, alcohol consumption, and obesity [22]. The platelet counts in the BC group showed no significant change compared to the control group (p>0.05). However, tumor markers such as Alpha-fetoprotein (AFP) (p=0.004), carcinoembryonic antigen (CEA) (p<0.001), CA19-9 (p<0.001), and liver function indicators including AST, total bilirubin, direct bilirubin, and their ratios were significantly elevated (p<0.001), while albumin showed a significant decrease (p=0.008). This is closely associated with liver function impairment in patients with bile duct cancer. In addition, the coagulation index D-dimer was also significantly elevated (p=0.025), which showed the hypercoagulable state of blood in cancer patients. However, other parameters such as ALT and PT did not differ between enrollments (p>0.05).

Table 1:

The general clinical baseline features of the participants.

Characteristics BC BS NC p-Value
n 18 27 15
Sex, n, % 0.340
Female 7 (11.7 %) 16 (26.7 %) 9 (15 %)
Male 11 (18.3 %) 11 (18.3 %) 6 (10 %)
Age, mean ± SD 68.778 ± 8.250 61.704 ± 13.076 61.800 ± 12.935 0.058
Platelet count (first admission) (125–350), mean ± SD 181.17 ± 65.51 204.85 ± 61.45 215.87 ± 41.71 0.217
AFP (<20), median (IQR) 3.4 (2.7, 4.7) 2.5 (1.9, 3.2) 3.4 (2.9, 4.6) 0.004 b
CEA (<5), median (IQR) 4.1 (2.6, 5.6) 1.8 (1.1, 2.4) 1.8 (1.3, 2.4) <0.001 c
CA19-9 (<25), median (IQR) 210.1 (69.7, 1814.6) 22.1 (5.5, 102.1) 9.3 (3.4, 19.0) <0.001 c
Smoke, n, % 0.089
Yes 8 (13.3 %) 4 (6.7 %) 4 (6.7 %)
No 10 (16.7 %) 23 (38.3 %) 11 (18.3 %)
Drink, n, % 0.024 a
No 12 (20.0 %) 26 (43.3 %) 13 (21.7 %)
Yes 6 (10.0 %) 1 (1.7 %) 2 (3.3 %)
ALB (40–55), mean ± SD 34.072 ± 4.818 36.693 ± 6.208 40.200 ± 4.545 0.008 b
ALT (9−50), median (IQR) 100.1 (58.9, 162.3) 65.1 (19.7, 344.6) 26.6 (14.9, 91.4) 0.061
AST (15–40), median (IQR) 82.5 (54.2, 142.8) 45.3 (20.2, 138.6) 21.8 (19.6, 31.3) <0.001 c
HBV infection, n, % 0.621
No 16 (26.7 %) 26 (43.3 %) 14 (23.3 %)
Yes 2 (3.3 %) 1 (1.7 %) 1 (1.7 %)
HCV infection, n, % 0.489
No 16 (26.7 %) 26 (43.3 %) 13 (21.7 %)
Yes 2 (3.3 %) 1 (1.7 %) 2 (3.3 %)
TB (<26.0), median (IQR) 255.5 (51.3, 394.8) 30.0 (17.1, 77.7) 11.8 (8.8, 20.5) <0.001 c
DBIL (<4.0), median (IQR) 138.2 (26.7, 287.6) 12.5 (3.1, 59.6) 2.5 (1.7, 4.9) <0.001 c
DBIL/TB, median (IQR) 0.579 (0.526, 0.733) 0.398 (0.248, 0.588) 0.211 (0.181, 0.273) <0.001 c
PT (9.9–12.8), median (IQR) 11.9 (11.5, 12.3) 11.6 (11.0, 12.4) 11.2 (10.8, 12.2) 0.350
D-dimer (0–243), median (IQR) 273.5 (153.8, 779.5) 144.0 (114.0, 377.0) 134.0 (84.5, 184.0) 0.025 a

  1. BC, bile duct cancer; BS, bile duct stone; NC, nature control; AFP, alpha-fetoprotein; CEA, carcinoembryonic antigen; ALB, albumin; ALT, alanine transaminase; AST, aspartate aminotransferase; HBV, hepatitis B virus; HCV, hepatitis C virus; TB, total bilirubin; DBIL, direct bilirubin; PT, Prothrombin time; SD, standard deviation. Bolded p-Values indicate significant differences, ap<0.05, bp<0.01, cp<0.001.

The relative expression levels of each candidate gene in platelets in BC, BS, and NC

We detected the change in expression levels of candidate genes in platelets between CCA and control groups by qRT-PCR assays (Figure 5). The relative expression levels of genes METAP1, SEC11A, WDR82, RASGRP1, and EED in platelets showed significant differences in bile duct cancer compared to the other two control groups (p<0.05), and all of them exhibited a downward trend, consistent with the bioinformatics prediction results. However, there was no significant change in values between the two control groups, contrary to previous studies [23]. Taking a comprehensive view, it is likely that the unavoidable bias observed in this study may be due to the limited sample size and the source of the samples. The expression levels of the CRYM gene showed a significant difference between BC and NC (p<0.001), but there was no significant difference compared to patients with BS. The expression level of the IFI27 gene showed no significant difference among the samples in all groups. However, the trends in the expression levels of these two genes are consistent with the previous bioinformatics prediction results. In summary, the experimental results have validated the predictions of bioinformatics analysis and confirmed that these seven significantly differentially expressed genes identified possess the capability for diagnosis and early diagnosis of BC.

Figure 5: Analysis of the expression levels of candidate genes in BC, BS, and NC (A–G) are the relative differential expression of genes (A) CRYM (B) IFI27 (C) METAP1 (D) SEC11A (E) WDR82 (F) RASGRP1, and (G) EED in bile duct cancer, bile duct stones, and normal controls, respectively. ns, *, **, *** represent p>0.05, p<0.05, p<0.01, p<0.001, respectively.
Figure 5:

Analysis of the expression levels of candidate genes in BC, BS, and NC (A–G) are the relative differential expression of genes (A) CRYM (B) IFI27 (C) METAP1 (D) SEC11A (E) WDR82 (F) RASGRP1, and (G) EED in bile duct cancer, bile duct stones, and normal controls, respectively. ns, *, **, *** represent p>0.05, p<0.05, p<0.01, p<0.001, respectively.

TEP mRNA as a novel biomarker for the diagnosis of cholangiocarcinoma (CCA)

To assess the diagnostic efficiency of candidate TEP mRNAs for cholangiocarcinoma, the ROC curves were calculated. As shown in Figure 6, except for the IFI27 gene, the other six genes, METAP1, SEC11A, WDR82, RASGRP1, EED, and CRYM, showed significant diagnostic efficacy (AUC>0.7) in BC and NC. Among them, WDR82 and CRYM demonstrated exceptionally significant diagnostic significance, with AUC values of 0.939 and 0.942, respectively. Their sensitivity, specificity, and accuracy were 0.895, 0.846, and 0.867 for WDR82, and 0.933, 0.933, and 0.933 for CRYM, respectively. These results were significantly better than conventional digestive tract tumor markers such as AFP (AUC of 0.556) (Supplementary Figure S2A), CEA (AUC of 0.833) (Supplementary Figure S2B), and CA19-9 (AUC of 0.889). Therefore, it can be concluded that these six platelet genes possess strong discriminatory ability for BC and have the potential to become novel and reliable biomarkers, which are of great importance in guiding the early diagnosis and intervention of CCA. It is worth mentioning that the AUC, sensitivity, and accuracy of constructing an ROC curve (Figure 6I) by combining the seven candidate genes were 0.806, 1.000, and 0.833, respectively. It can be seen that this model has reliable value for the diagnosis and early diagnosis of CCA. Although its diagnostic capability did not show a significant advantage over CA19-9 in the experimental data, this may be attributed to the limited sample size. In subsequent studies, we will continue to increase the sample size, improve relevant experimental data, and further assess the rationality and applicability of the model construction.

Figure 6: Analysis of the receiver operating characteristics (ROC) curves for the most significant differentiating genes (A–I) ROC curves and AUC values for the genes (A) CRYM (B) IFI27 (C) METAP1 (D) SEC11A (E) WDR82 (F) RASGRP1 (G) EED (H)CA19-9, and (I) the combined respectively, as well as sensitivity, specificity, and accuracy data.
Figure 6:

Analysis of the receiver operating characteristics (ROC) curves for the most significant differentiating genes (A–I) ROC curves and AUC values for the genes (A) CRYM (B) IFI27 (C) METAP1 (D) SEC11A (E) WDR82 (F) RASGRP1 (G) EED (H)CA19-9, and (I) the combined respectively, as well as sensitivity, specificity, and accuracy data.

Changes in the expression of each candidate gene in vitro experiments

We validated changes in the expression levels of candidate genes in the in vitro experiments using the CCA cell line HUCC-T1 and the liver bile duct cell line THLE-3. It turns out that the results demonstrated that the genes CRYM, IFI27, METAP1, SEC11A, WDR82, RASGRP1, and EED all exhibit changing trends consistent with bioinformatics and clinical samples (Figure 7). With the exception of SEC11A and RASGRP1, the remaining five genes have obvious significance (p<0.05). Therefore, cell experiments further verified the bioinformatics predictions and experimental results of clinical samples, confirming that TEP RNAs has the ability to diagnose and early diagnose CCA.

Figure 7: Analysis of the most significant differential genes in the cholangiocarcinoma (CCA) cell line HUCC-T1 and the hepatocholangial-tube cell line THLE-3 (A–G) are the relative differential expression of genes (A) CRYM (B) IFI27 (C) METAP1 (D) SEC11A (E) WDR82 (F) RASGRP1, and (G) EED in CCA cells and normal hepatocholangiocyte (THLE-3), respectively. *, **, *** represent p<0.05, p<0.01, p<0.001, respectively.
Figure 7:

Analysis of the most significant differential genes in the cholangiocarcinoma (CCA) cell line HUCC-T1 and the hepatocholangial-tube cell line THLE-3 (A–G) are the relative differential expression of genes (A) CRYM (B) IFI27 (C) METAP1 (D) SEC11A (E) WDR82 (F) RASGRP1, and (G) EED in CCA cells and normal hepatocholangiocyte (THLE-3), respectively. *, **, *** represent p<0.05, p<0.01, p<0.001, respectively.

Discussion

CCA is a malignant tumor that originates from the epithelial cells of the bile ducts, accounting for 3 % of all digestive tract tumors [24]. Due to its insidious onset and atypical early clinical symptoms, CCA is often diagnosed at an advanced stage when patients seek medical attention. As a result, the overall prognosis is poor, with a five-year survival rate of less than 5 % [25]. Currently, the diagnosis of CCA relies primarily on imaging and further cytology or histologic confirmation [26], 27]. The tumor marker carbohydrate antigen 19–9 (CA19-9) is currently the only liquid biopsy tool used clinically to aid in the diagnosis of CCA, but its diagnostic ability is low, especially in the early stages of CCA [6]. Therefore, there is an urgent need to discover a new diagnostic method to be able to detect CCA at an early stage, improve its early diagnosis rate, and reduce the possibility of mid- and late-stage development and mortality of patients. The use of blood-based liquid biopsy as a novel, noninvasive biomarker detection method in recent years has led to revolutionary advances in clinical diagnostic and treatment technologies. By analyzing the presence of bioactive substances and RNA in plasma or serum, liquid biopsy can provide a wealth of diagnostic and testing information for clinical purposes, including early cancer screening, tumor diagnosis and monitoring, and assessment of treatment response. Increasing evidence has revealed multiple roles of TEP mRNAs in various pathophysiological processes and cancer pathogenesis [20]. TEPs can respond to tumor growth and treatment and recognize specific tumor types [28].

In this study, two CCA TEP mRNA datasets were mined from the GEO database, and bioinformatics analysis was performed to screen and identify seven genes that were most significantly differentially expressed in CCA patients compared to healthy controls, which were CRYM, IFI27, METAP1, SEC11A, WDR82, RASGRP1, and EED, and jointly constructed a CCA multigene diagnostic model that was validated to be of high accuracy at the bioinformatics level. These are genes that can encode proteins.

CRYM gene expression generates a crucial intracellular high-affinity T3-binding protein, which balances the equilibrium of free T3 in the cytoplasm, consequently modulating its effects. Its deficiency leads to increased thyroid hormone activity in affected cells. The potential oncogenic or tumor-suppressive functions of CRYM have not been extensively studied [29]. However, there is growing evidence that CRYM plays a significant role in tumorigenesis by both upregulating and downregulating its expression. In non-small cell lung cancer (NSCLC), CRYM expression is found to be twice as high compared to matched normal tissue [30]; Additionally, gene expression profiling in primary and metastatic leiomyosarcoma (LMS), an aggressive uterine tumor, reveals significant CRYM expression in metastatic LMS, which may provide insights into the tumor progression of this cancer [31]. Recently, CRYM has been identified as one of seven candidate genes that enhance hepatitis C virus replication in liver cancer cells, potentially contributing to advanced liver cancer associated with this disease [32].

IFI27 is implicated in the apoptosis signaling pathway, and its enhanced stability promotes angiogenesis and malignant progression in esophageal squamous cell carcinoma [33]. Additionally, it has been observed to be highly expressed in breast cancer [34]. In oral and tongue squamous cell carcinoma, downregulation of IFI27 can inhibit cell proliferation and invasion while promoting apoptosis [35]. Furthermore, IFI27 has been identified as a potential prognostic biomarker for pancreatic cancer [36].

METAP1 is mainly involved in platelet aggregation. Research has revealed that knocking down METAP1 impairs MTT viability and diminishes competitive cell growth in mouse breast cancer cells [37].

SEC11A plays a role in promoting tumor cell invasion and migration in bladder cancer [38], gastric cancer [39], and head and neck squamous cell carcinoma (HNSCC) [40]. Upregulation of SEC11A can serve as a biomarker for poor prognosis in cancer patients [41]. WDR82 is inversely correlated with the development and progression of colorectal cancer [42], medulloblastoma [43], and lung cancer [44].

RASGRP1 deficiency has been linked to life-threatening immune dysregulation, severe autoimmune manifestations, and increased susceptibility to EBV-induced B-cell malignancies [45]. RASGRP1 limits intestinal epithelial cell growth by counteracting proliferative EGFR-SOS1-Ras signaling [46]. In hepatocellular carcinoma (HCC), upregulation of RASGRP1 is associated with poor prognosis, highlighting its potential as a novel biomarker and therapeutic target [47]. Additionally, under the cooperative effects of Nras Q61R/+ and Kras−/−, downregulation of RASGRP1 promotes lymphoid-myeloid leukemia in early T-cell progenitors, indicating that RASGRP1 serves as a negative regulator of Ras/ERK signaling in oncogenic Nras-driven ETP-like leukemia [48].

EED, as one of the three subunits of the polycomb repressor complex 2 (PRC2), plays a role in the development of cholangiocarcinoma by promoting DNA methylation at gene promoters, thereby influencing patient prognosis and response [49]. Moreover, recent studies suggest that EED may be involved in the metabolic processes of cervical cancer [50]. Our findings could have significant implications for understanding cholangiocarcinogenesis and its progression.

The clinical data show that the experimental and control groups were well-matched in terms of gender, age distribution, smoking status, and the presence of hepatitis virus infection, ensuring objectivity. However, due to the limitations of the study, it is not yet possible to determine with certainty whether these factors significantly influence the development of cholangiocarcinoma. On the other hand, the elevated levels of several biochemical markers clearly demonstrate significant differences between the experimental and control groups. The results of the experiments demonstrated significant differences in the expression of the genes METAP1, SEC11A, WDR82, RASGRP1, and EED in platelets between the CCA group and both control groups. While the expression of the CRYM gene was markedly higher in CCA patients compared to healthy individuals, it did not exhibit significant differences when compared to patients with bile duct stones. The expression levels of the IFI27 gene showed no significant differences among the various sample groups. In summary, the factors that appear to make the experimental results inconsistent with expectations are primarily attributed to the inevitable bias stemming from the limited sample size and the single source of the samples.

ROC curve analysis showed that six genes, METAP1, SEC11A, WDR82, RASGRP1, EED, and CRYM, apart from the IFI27 gene, exhibited significant diagnostic potential for distinguishing CCA from healthy populations (AUC>0.7). Among them, WDR82 and CRYM exhibited notably substantial diagnostic significance, significantly superior to conventional gastrointestinal tumor markers such as AFP, CEA, and CA19-9. Therefore, it can be concluded that these six TEP mRNAs possess strong discriminatory ability for CCA and are promising to become clinical auxiliary diagnostic biomarkers for CCA with potential application value. On this basis, ROC curve analysis of the model jointly constructed by the seven candidate genes revealed an AUC of 0.806, sensitivity of 1.000, and accuracy of 0.833. However, its diagnostic performance did not exhibit a significant advantage over that of CA19-9, which was considered to be related to the insufficient sample size. To further judge the rationality and applicability of model construction, we intend to expand the sample size and improve relevant experimental data in future investigations. The results of the vitro experiments further reinforced the reliability of the predictions made through bioinformatics and the experimental findings from clinical samples. In conclusion, our study provides new ideas for the early diagnosis of CCA, and all selected biomarkers can be successfully detected from the sample. While the diagnostic efficiency of this model was found to be slightly lower than that of CA19-9, given the limitations of the experiment, this does not diminish the strong potential of platelet-based liquid biopsy technology as a reliable, accurate, and widely applicable tool for early diagnosis, personalized treatment, and better prognostic assessment of cancer patients [51].

This study demonstrates that combining transcriptomics with clinical information can help improve future research and potential applications for clinical diagnosis. Inevitably, our study has some limitations. Firstly, gene expression profiles may be affected by confounding factors, such as individual variability between samples, small sample sizes, and platforms. Therefore, more microarray data and stricter quality control need to be added to minimize errors. Secondly, a portion of the data used in our analysis was sourced from online databases. Thirdly, both our CCA patients and the controls are from northern China, which may introduce some bias into the experimental results. Additionally, this is a single-center study with a relatively small clinical sample size, as only 60 peripheral blood samples were collected in the clinic. Our experiments provide preliminary evidence for TEP RNA as a biomarker for the diagnosis of CCA, but there is still a need to expand the sample size and conduct multicenter, large-sample randomized controlled studies and biological investigations at a later stage. Fourthly, we did not investigate the potential mechanisms by which the candidate genes may play a role in CCA. Future studies will explore the specific interactions between the two in greater depth. More importantly, our study did not analyze the correlation between TEP mRNA levels and tumor characteristics such as tumor stage, microvascular invasion, and differentiation in CCA patients. In conclusion, based on limited preliminary data, we found that TEP RNA has the potential to be a novel and reliable biomarker for the diagnosis of CCA.

Conclusions

Through bioinformatics methods, we obtained seven hub genes that connect platelets and CCA and explored their feasibility as biomarkers for CCA. Based on Lasso regression analysis, a diagnostic model was constructed that can diagnose CCA patients by detecting the expression of multiple genes in platelets. While our findings do not yet confirm the model’s clinical utility with absolute certainty, they suggest that platelet RNA profiles may be applicable for the diagnosis and early detection of CCA. Moving forward, further large-sample, multicenter, and prospective studies are needed to validate the diagnostic significance of TEP mRNA profiling and expand its application to various aspects of cancer management, including monitoring, prognosis, and recurrence detection.

Corresponding authors: Jiefeng He and Haoliang Zhao, Department of Hepatobiliary Surgery, Shanxi Bethune Hospital, Shanxi Academy of Medical Sciences, Tongji Shanxi Hospital, Taiyuan, 030032, People’s Republic of China, E-mail: (J. He), (H. Zhao)

Funding source: Shanxi ‘136’ Leading Clinical Key Specialty

Award Identifier / Grant number: No. 2019XY002

Funding source: the National Natural Science Foundation of China

Award Identifier / Grant number: No. 82073090

Funding source: Shanxi Provincial Key Laboratory of Hepatobiliary and Pancreatic Diseases

Funding source: Research Project Supported by Shanxi Scholarship Council of China

Award Identifier / Grant number: No. 2021-116

Funding source: Human Resources and Social Security Department System of Shanxi Province

Award Identifier / Grant number: No.20210001

  1. Research ethics: This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Medical and Health Ethics Review Committee of Shanxi Bethune Hospital (No. YXLL-2023-170). Informed consent was obtained from all individual participants included in the study.

  2. Informed consent: Informed consent was obtained from all individuals included in this study, or their legal guardians or wards.

  3. Author contributions: All authors listed in this article participated in the creation of the article. Haiyang Hu led the creation of the manuscript and participated in the production of tables and figures, and Haoliang Zhao and Jiefeng He participated in the planning and final review of the thesis theme. All authors read and approved the final manuscript.

  4. Use of Large Language Models, AI and Machine Learning Tools: None declared.

  5. Conflict of interest: The authors have no relevant financial or non-financial interests to disclose.

  6. Research funding: This article was supported by the National Natural Science Foundation of China (Grant No. 82073090), Fund Program for the Scientific Activities of Selected Returned Overseas Professionals in Shanxi Province (Grant No.20210001), Research Project Supported by Shanxi Scholarship Council of China (Grant No. 2021–116), Shanxi ‘136’ Leading Clinical Key Specialty (Grant No. 2019XY002), and Shanxi Provincial Key Laboratory of Hepatobiliary and Pancreatic Diseases.

  7. Data availability: The raw data can be obtained on request from the corresponding author.

References

1. Li, Y, Yu, J, Zhang, Y, Peng, C, Song, Y, Liu, S. Advances in targeted therapy of cholangiocarcinoma. Ann Med 2024;56. https://doi.org/10.1080/07853890.2024.2310196.Search in Google Scholar PubMed PubMed Central

2. Shah, UJ, Alsulimani, A, Ahmad, F, Mathkor, DM, Alsaieedi, A, Harakeh, S, et al.. Bioplatforms in liquid biopsy: advances in the techniques for isolation, characterization and clinical applications. Biotechnol Genet Eng Rev 2022;38:339–83. https://doi.org/10.1080/02648725.2022.2108994.Search in Google Scholar PubMed

3. Eslami, SZ, Cortes-Hernandez, LE, Cayrefourcq, L, Alix-Panabieres, C. The different facets of liquid biopsy: a kaleidoscopic view. Cold Spring Harb Perspect Med 2020;1. https://doi.org/10.1101/cshperspect.a037333.Search in Google Scholar PubMed PubMed Central

4. Siravegna, G, Marsoni, S, Siena, S, Bardelli, A. Integrating liquid biopsies into the management of cancer. Nat Rev Clin Oncol 2017;14:531–48. https://doi.org/10.1038/nrclinonc.2017.14.Search in Google Scholar PubMed

5. Ilie, M, Hofman, P. Pros: can tissue biopsy be replaced by liquid biopsy? Transl Lung Cancer Res 2016;5:420–3. https://doi.org/10.21037/tlcr.2016.08.06.Search in Google Scholar PubMed PubMed Central

6. Lapitz, A, Azkargorta, M, Milkiewicz, P, Olaizola, P, Zhuravleva, E, Grimsrud, MM, et al.. Liquid biopsy-based protein biomarkers for risk prediction, early diagnosis, and prognostication of cholangiocarcinoma. J Hepatol 2023;79:93–108. https://doi.org/10.1016/j.jhep.2023.02.027.Search in Google Scholar PubMed PubMed Central

7. Banales, JM, Marin, J, Lamarca, A, Rodrigues, PM, Khan, SA, Roberts, LR, et al.. Cholangiocarcinoma 2020: the next horizon in mechanisms and management. Nat Rev Gastroenterol Hepatol 2020;17:557–88. https://doi.org/10.1038/s41575-020-0310-z.Search in Google Scholar PubMed PubMed Central

8. Cossu, AM, Scrima, M, Lombardi, A, Grimaldi, A, Russo, M, Ottaiano, A, et al.. Future directions and management of liquid biopsy in non-small cell lung cancer. Explor Target Antitumor Ther 2020;1:239–52. https://doi.org/10.37349/etat.2020.00015.Search in Google Scholar PubMed PubMed Central

9. Rodríguez, J, Avila, J, Rolfo, C, Ruíz-Patiño, A, Russo, A, Ricaurte, L, et al.. When tissue is an issue the liquid biopsy is nonissue: a review. Oncol Ther 2021;9:89–110. https://doi.org/10.1007/s40487-021-00144-6.Search in Google Scholar PubMed PubMed Central

10. Ma, C, Fu, Q, Diggs, LP, Mcvey, JC, Mccallen, J, Wabitsch, S, et al.. Platelets control liver tumor growth through P2Y12-dependent CD40L release in NAFLD. Cancer Cell 2022;40:986–98. https://doi.org/10.1016/j.ccell.2022.08.004.Search in Google Scholar PubMed PubMed Central

11. Varkey, J, Nicolaides, T. Tumor-educated platelets: a review of current and potential applications in solid tumors. Cureus 2021;13:e19189. https://doi.org/10.7759/cureus.19189.Search in Google Scholar PubMed PubMed Central

12. Heeke, S, Mograbi, B, Alix-Panabieres, C, Hofman, P. Never travel alone: the crosstalk of circulating tumor cells and the blood microenvironment. Cells 2019;8:714. https://doi.org/10.3390/cells8070714.Search in Google Scholar PubMed PubMed Central

13. Jiang, X, Wong, K, Khankhel, AH, Zeinali, M, Reategui, E, Phillips, MJ, et al.. Microfluidic isolation of platelet-covered circulating tumor cells. Lab Chip 2017;17:3498–503. https://doi.org/10.1039/c7lc00654c.Search in Google Scholar PubMed PubMed Central

14. Wang, L, Wang, X, Guo, E, Mao, X, Miao, S. Emerging roles of platelets in cancer biology and their potential as therapeutic targets. Front Oncol 2022;12. https://doi.org/10.3389/fonc.2022.939089.Search in Google Scholar PubMed PubMed Central

15. Best, MG, Wesseling, P, Wurdinger, T. Tumor-educated platelets as a noninvasive biomarker source for cancer detection and progression monitoring. Cancer Res 2018;78:3407–12. https://doi.org/10.1158/0008-5472.can-18-0887.Search in Google Scholar PubMed

16. Ritchie, ME, Phipson, B, Wu, D, Hu, Y, Law, CW, Shi, W, et al.. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43:e47. https://doi.org/10.1093/nar/gkv007.Search in Google Scholar PubMed PubMed Central

17. Yu, G, Wang, L, Han, Y, He, Q. ClusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 2012;16:284–7. https://doi.org/10.1089/omi.2011.0118.Search in Google Scholar PubMed PubMed Central

18. Sheppard, CA, Allen, RC, Austin, GE, Young, AN, Ribeiro, MA, Fantz, CR. Paraprotein interference in automated chemistry analyzers. Clin Chem 2005;51:1077–8. https://doi.org/10.1373/clinchem.2004.045740.Search in Google Scholar PubMed

19. Wen, X, Yang, G, Dong, Y, Luo, L, Cao, B, Mengesha, BA, et al.. Selection and validation of reference genes for pan-cancer in platelets based on RNA-sequence data. Front Genet 2022;13. https://doi.org/10.3389/fgene.2022.913886.Search in Google Scholar PubMed PubMed Central

20. Best, MG, Sol, N, Kooi, I, Tannous, J, Westerman, BA, Rustenburg, F, et al.. RNA-seq of tumor-educated platelets enables blood-based pan-cancer, multiclass, and molecular pathway cancer diagnostics. Cancer Cell 2015;28:666–76. https://doi.org/10.1016/j.ccell.2015.09.018.Search in Google Scholar PubMed PubMed Central

21. Taniai, M. Alcohol and hepatocarcinogenesis. Clin Mol Hepatol 2020;26:736–41. https://doi.org/10.3350/cmh.2020.0203.Search in Google Scholar PubMed PubMed Central

22. Cho, IR, Yi, SW, Choi, JS, Yi, JJ. Comparison of risk factors for cholangiocarcinoma and hepatocellular carcinoma: a prospective cohort study in Korean adults. Cancers 2022;14:1709. https://doi.org/10.3390/cancers14071709.Search in Google Scholar PubMed PubMed Central

23. Rebholz, C, Krawczyk, M, Lammert, F. Genetics of gallstone disease. Eur J Clin Invest 2018;48:e12935. https://doi.org/10.1111/eci.12935.Search in Google Scholar PubMed

24. Benavides, M, Anton, A, Gallego, J, Gomez, MA, Jimenez-Gordo, A, La Casta, A, et al.. Biliary tract cancers: SEOM clinical guidelines. Clin Transl Oncol 2015;17:982–7. https://doi.org/10.1007/s12094-015-1436-2.Search in Google Scholar PubMed PubMed Central

25. Feng, L, Guangwen, Z. Hilar cholangiocarcinoma: a meta-analysis of 2280 cases. Chin J Hepatobiliary Surg 2013;12:171–6.Search in Google Scholar

26. Izquierdo-Sanchez, L, Lamarca, A, La Casta, A, Buettner, S, Utpatel, K, Klumpen, HJ, et al.. Cholangiocarcinoma landscape in Europe: diagnostic, prognostic and therapeutic insights from the ENSCCA Registry. J Hepatol 2022;76:1109–21. https://doi.org/10.1016/j.jhep.2021.12.010.Search in Google Scholar PubMed

27. Macias, R, Cardinale, V, Kendall, TJ, Avila, MA, Guido, M, Coulouarn, C, et al.. Clinical relevance of biomarkers in cholangiocarcinoma: critical revision and future directions. Gut 2022;71:1669–83. https://doi.org/10.1136/gutjnl-2022-327099.Search in Google Scholar PubMed

28. Joosse, SA, Pantel, K. Tumor-educated platelets as liquid biopsy in cancer patients. Cancer Cell 2015;28:552–4. https://doi.org/10.1016/j.ccell.2015.10.007.Search in Google Scholar PubMed

29. Aksoy, O, Hantusch, B, Kenner, L. Emerging role of T3-binding protein mu-crystallin (CRYM) in health and disease. Trends Endocrinol Metab 2022;33:804–16. https://doi.org/10.1016/j.tem.2022.09.003.Search in Google Scholar PubMed

30. Chong, IW, Chang, MY, Chang, HC, Yu, YP, Sheu, CC, Tsai, JR, et al.. Great potential of a panel of multiple hMTH1, SPD, ITGA11 and COL11A1 markers for diagnosis of patients with non-small cell lung cancer. Oncol Rep 2006;16:981–8. https://doi.org/10.3892/or.16.5.981.Search in Google Scholar

31. Davidson, B, Abeler, VM, Forsund, M, Holth, A, Yang, Y, Kobayashi, Y, et al.. Gene expression signatures of primary and metastatic uterine leiomyosarcoma. Hum Pathol 2014;45:691–700. https://doi.org/10.1016/j.humpath.2013.11.003.Search in Google Scholar PubMed PubMed Central

32. Dachert, C, Gladilin, E, Binder, M. Gene expression profiling of different Huh7 variants reveals novel hepatitis C virus host factors. Viruses 2019;12:36. https://doi.org/10.3390/v12010036.Search in Google Scholar PubMed PubMed Central

33. Zhang, Y, Chen, C, Liu, Z, Guo, H, Lu, W, Hu, W, et al.. PABPC1-induced stabilization of IFI27 mRNA promotes angiogenesis and malignant progression in esophageal squamous cell carcinoma through exosomal miRNA-21-5p. J Exp Clin Cancer Res 2022;41:111. https://doi.org/10.1186/s13046-022-02339-9.Search in Google Scholar PubMed PubMed Central

34. Cervantes-Badillo, MG, Paredes-Villa, A, Gomez-Romero, V, Cervantes-Roldan, R, Arias-Romero, LE, Villamar-Cruz, O, et al.. IFI27/ISG12 downregulates estrogen receptor alpha transactivation by facilitating its interaction with CRM1/XPO1 in breast cancer cells. Front Endocrinol 2020;11. https://doi.org/10.3389/fendo.2020.568375.Search in Google Scholar PubMed PubMed Central

35. Wang, H, Qiu, X, Lin, S, Chen, X, Wang, T, Liao, T. Knockdown of IFI27 inhibits cell proliferation and invasion in oral squamous cell carcinoma. World J Surg Oncol 2018;16:64. https://doi.org/10.1186/s12957-018-1371-0.Search in Google Scholar PubMed PubMed Central

36. Huang, S, Zhao, J, Song, J, Li, Y, Zuo, R, Sa, Y, et al.. Interferon alpha-inducible protein 27 (IFI27) is a prognostic marker for pancreatic cancer based on comprehensive bioinformatics analysis. Bioengineered 2021;12:8515–28. https://doi.org/10.1080/21655979.2021.1985858.Search in Google Scholar PubMed PubMed Central

37. Holzen, L, Mitschke, J, Schonichen, C, Hess, ME, Ehrenfeld, S, Boerries, M, et al.. RNA interference screens discover proteases as synthetic lethal partners of PI3K inhibition in breast cancer cells. Theranostics 2022;12:4348–73. https://doi.org/10.7150/thno.68299.Search in Google Scholar PubMed PubMed Central

38. Shigematsu, Y, Oue, N, Sekino, Y, Sakamoto, N, Sentani, K, Uraoka, N, et al.. SEC11A expression is associated with basal-like bladder cancer and predicts patient survival. Pathobiology 2019;86:208–16. https://doi.org/10.1159/000497206.Search in Google Scholar PubMed

39. Suematsu, H, Sakamaki, K, Oue, N, Hiroshima, Y, Kimura, Y, Onuma, S, et al.. Clinical significance of SEC11A expression in patients with locally advanced gastric cancer. Anticancer Res 2022;42:5885–90. https://doi.org/10.21873/anticanres.16097.Search in Google Scholar PubMed

40. Shen, H, Yi, F, Ding, Z, Liu, W, Liu, P, Wang, Z, et al.. SEC11A contributes to tumour progression of head and neck squamous cell carcinoma. Heliyon 2023;9. https://doi.org/10.1016/j.heliyon.2023.e14958.Search in Google Scholar PubMed PubMed Central

41. Hu, C, Fan, J, He, G, Dong, C, Zhou, S, Zheng, Y. Signal peptidase complex catalytic subunit SEC11A upregulation is a biomarker of poor prognosis in patients with head and neck squamous cell carcinoma. PLoS One 2022;17:e269166. https://doi.org/10.1371/journal.pone.0269166.Search in Google Scholar PubMed PubMed Central

42. Li, P, Weng, Z, Li, P, Hu, F, Zhang, Y, Guo, Z, et al.. BATF3 promotes malignant phenotype of colorectal cancer through the S1PR1/p-STAT3/miR-155-3p/WDR82 axis. Cancer Gene Ther 2021;28:400–12. https://doi.org/10.1038/s41417-020-00223-2.Search in Google Scholar PubMed

43. Song, L, Luan, B, Xu, Q, Shi, R, Wang, X. microRNA-155-3p delivered by M2 macrophages-derived exosomes enhances the progression of medulloblastoma through regulation of WDR82. J Transl Med 2022;20:13. https://doi.org/10.1186/s12967-021-03156-y.Search in Google Scholar PubMed PubMed Central

44. Lei, J, Chen, P, Zhang, F, Zhang, N, Zhu, J, Wang, X, et al.. M2 macrophages-derived exosomal microRNA-501-3p promotes the progression of lung cancer via targeting WD repeat domain 82. Cancer Cel Int 2021;21:91. https://doi.org/10.1186/s12935-021-01783-5.Search in Google Scholar PubMed PubMed Central

45. Somekh, I, Marquardt, B, Liu, Y, Rohlfs, M, Hollizeck, S, Karakukcu, M, et al.. Novel mutations in RASGRP1 are associated with immunodeficiency, immune dysregulation, and EBV-induced lymphoma. J Clin Immunol 2018;38:699–710. https://doi.org/10.1007/s10875-018-0533-8.Search in Google Scholar PubMed

46. Depeille, P, Henricks, LM, Van De Ven, RA, Lemmens, E, Wang, CY, Matli, M, et al.. RasGRP1 opposes proliferative EGFR-SOS1-Ras signals and restricts intestinal epithelial cell growth. Nat Cel Biol 2015;17:804–15. https://doi.org/10.1038/ncb3175.Search in Google Scholar PubMed PubMed Central

47. Zhang, X, Liu, Y, Yang, R, Guo, Y, Yan, M, Xiao, Y, et al.. Phosphorylation of RasGRP1 by Shc3 prevents RasGRP1 degradation and contributes to Ras/c-Jun activation in hepatocellular carcinoma. Mol Cel Biochem 2024;479:2307–21. https://doi.org/10.1007/s11010-023-04839-4.Search in Google Scholar PubMed

48. Wen, Z, Yun, G, Hebert, A, Kong, G, Ranheim, EA, Finn, R, et al.. Nras Q61R/+ and Kras–/– cooperate to downregulate rasgrp1 and promote lympho-myeloid leukemia in early T-cell precursors. Blood 2021;137:3259–71. https://doi.org/10.1182/blood.2020009082.Search in Google Scholar PubMed PubMed Central

49. Wu, G, Wang, Q, Wang, D, Xiong, F, Liu, W, Chen, J, et al.. Targeting polycomb repressor complex 2-mediated bivalent promoter epigenetic silencing of secreted frizzled-related protein 1 inhibits cholangiocarcinoma progression. Clin Transl Med 2023;13:e1502. https://doi.org/10.1002/ctm2.1502.Search in Google Scholar PubMed PubMed Central

50. Xi, Y, Zheng, P, Xi, W, Fu, T. Exploring the impact of estrogenic endocrine disruptors on cervical cancer progression: a transcriptome analysis and prognostic model development. Ecotoxicol Environ Saf 2024;285. https://doi.org/10.1016/j.ecoenv.2024.117025.Search in Google Scholar PubMed

51. Hu, H, Song, H, Han, B, Zhao, H, He, J. Tumor-educated platelet RNA and circulating free RNA: emerging liquid biopsy markers for different tumor types. Front Biosci 2024;29:80. https://doi.org/10.31083/j.fbl2902080.Search in Google Scholar PubMed

Supplementary Material

This article contains supplementary material (https://doi.org/10.1515/oncologie-2024-0520).

Received: 2024-10-09
Accepted: 2024-12-25
Published Online: 2025-01-13
Published in Print: 2025-03-26

© 2024 the author(s), published by De Gruyter on behalf of Tech Science Press (TSP)

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Frontmatter
  2. Review Articles
  3. Targeting HER2 in metastatic urothelial carcinoma: a contemporary review
  4. Advances in research on the impact of exosomal miRNAs on liver metastasis of colorectal cancer
  5. Sebaceous carcinoma of the intraoral origin: a literature review
  6. Role of emerging theranostic technologies in precision oncology: revolutionizing cancer diagnosis and treatment
  7. Research Articles
  8. Half-body irradiation with dose escalation in the era of advanced systemic therapies: unveiling new therapeutic opportunities
  9. Characteristics of MMR protein expression in colorectal cancer and MMR gene variations in Vietnamese patients with Lynch syndrome
  10. BMPR1A promotes the proliferation of colorectal cancer cells through the activation of Smad1
  11. Construction and validation of a diagnostic model for cholangiocarcinoma based on tumor-educated platelet RNA expression profiles
  12. Prognostic implications of PCSK9 expression in HER2-positive breast cancer
  13. Rapid Communication
  14. Potential impact of hormone replacement therapy on the risk of hepatocellular carcinoma in women of the PLCO cohort
  15. Article Commentary
  16. “Plant-based and ketogenic diets as diverging paths to address cancer”: a commentary concerning the supposed superiority of a plant-based diet
  17. Novel insights into molecular landscape of advanced renal cell carcinoma
  18. Corrigendum
  19. Corrigendum to “Y27632 induces tongue squamous cell carcinoma cell apoptosis through MAPK-ERK/JNK signal”
https://doi.org/10.1515/oncologie-2024-0520
Keywords for this article
cholangiocarcinoma; liquid biopsy; tumor-educated platelet; diagnostic model RNA
Creative Commons
BY 4.0

Articles in the same Issue

  1. Frontmatter
  2. Review Articles
  3. Targeting HER2 in metastatic urothelial carcinoma: a contemporary review
  4. Advances in research on the impact of exosomal miRNAs on liver metastasis of colorectal cancer
  5. Sebaceous carcinoma of the intraoral origin: a literature review
  6. Role of emerging theranostic technologies in precision oncology: revolutionizing cancer diagnosis and treatment
  7. Research Articles
  8. Half-body irradiation with dose escalation in the era of advanced systemic therapies: unveiling new therapeutic opportunities
  9. Characteristics of MMR protein expression in colorectal cancer and MMR gene variations in Vietnamese patients with Lynch syndrome
  10. BMPR1A promotes the proliferation of colorectal cancer cells through the activation of Smad1
  11. Construction and validation of a diagnostic model for cholangiocarcinoma based on tumor-educated platelet RNA expression profiles
  12. Prognostic implications of PCSK9 expression in HER2-positive breast cancer
  13. Rapid Communication
  14. Potential impact of hormone replacement therapy on the risk of hepatocellular carcinoma in women of the PLCO cohort
  15. Article Commentary
  16. “Plant-based and ketogenic diets as diverging paths to address cancer”: a commentary concerning the supposed superiority of a plant-based diet
  17. Novel insights into molecular landscape of advanced renal cell carcinoma
  18. Corrigendum
  19. Corrigendum to “Y27632 induces tongue squamous cell carcinoma cell apoptosis through MAPK-ERK/JNK signal”
Sign up now to receive a 20% welcome discount
Institutional Access
How does access work?
Have an idea on how to improve our website?
Please write us.
© 2025 De Gruyter Brill
Downloaded on 25.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/oncologie-2024-0520/html
Scroll to top button