Novel serum marker combination for predicting colorectal cancer liver metastasis

Abstract

The invention relates to the field of biology, and discloses a novel serum marker combination for predicting colorectal cancer liver metastasis, the novel serum marker combination comprises a proteomic marker, an inflammation and metabolism dynamic marker, a genome and epigenetic marker, an exosome marker and a metabolite marker, the proteomic marker comprises Transferrin, Compleent C9 and Haptoglobin, and the metabolite marker comprises metabolite. The inflammation and metabolism dynamic markers comprise CRP (C-reactive protein) combined GPS (global positioning system) scores, lactic dehydrogenase and glutamate transpeptidase, and the exosome markers are exosomes secreted by tumor cells and comprise exosome nucleic acid and exosome protein. By fusing proteomics, inflammation and metabolic dynamics, genome and epiinheritance, exosome, metabolite and other markers, comprehensive analysis is performed from multiple dimensions of protein expression of cells, inflammation immune state, gene level change, intercellular communication, metabolic activity and the like.

Description

Technical Field

[0001] The present invention relates to the field of biotechnology, and particularly to a serum biomarker combination for predicting liver metastasis of colorectal cancer. Background Art

[0002] In the field of modern medicine, early diagnosis, precise treatment and prognosis evaluation of cancer have always been the key research directions. As one of the common malignant tumors worldwide, the incidence and mortality of colorectal cancer remain high, and liver metastasis is one of the main reasons for poor prognosis of colorectal cancer patients. With the continuous development of molecular biology, biochemistry and detection technologies, serum biomarker detection, as a non-invasive or minimally invasive detection method, has important application value in tumor diagnosis and disease monitoring due to its simple operation and strong repeatability.

[0003] The sensitivity and specificity of traditional serum biomarkers are limited. In patients with early liver metastasis of colorectal cancer, their levels may not increase significantly, which is likely to lead to missed diagnosis. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides a serum biomarker combination for predicting liver metastasis of colorectal cancer, which solves the problems that the sensitivity and specificity of traditional serum biomarkers are limited, and in patients with early liver metastasis of colorectal cancer, their levels may not increase significantly, easily leading to missed diagnosis.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A serum biomarker combination for predicting liver metastasis of colorectal cancer, including proteomic biomarkers, inflammation and metabolism dynamic biomarkers, genomic and epigenetic biomarkers, exosome biomarkers, metabolite biomarkers. The proteomic biomarkers include Transferrin, Complement C9, Haptoglobin. The inflammation and metabolism dynamic biomarkers include CRP combined with GPS score, lactate dehydrogenase, gamma-glutamyl transpeptidase. The genomic and epigenetic biomarkers include ctDNA mutation profile, miRNA, ctDNA methylation. The exosome biomarkers are exosomes secreted by tumor cells, including exosome nucleic acids and exosome proteins. The metabolite biomarkers include polyamines, amino acid metabolites, lipid metabolites.

[0006] Preferably, the CRP combined with GPS score is used to quantify the systemic inflammatory state.

[0007] Preferably, the ctDNA mutation profile includes APC, KRAS, BRAF.

[0008] Preferably, the miRNA includes miR-21, miR-31.

[0009] Preferably, the ctDNA methylation includes SEPT9 and RASSF1A.

[0010] Preferably, the exosomal nucleic acids include miR-122 and miR-155.

[0011] Preferably, the exosomal proteins include CD9, CD63, and EpCAM.

[0012] Preferably, the polyamines include putrescine, spermine, and spermidine.

[0013] Preferably, the amino acid metabolites include kynurenine and 5-hydroxytryptamine.

[0014] Preferably, the lipid metabolites include phosphatidylcholine and phosphatidylethanolamine.

[0015] The present invention provides a combination of serum markers for predicting liver metastasis of colorectal cancer, having the following beneficial effects: 1. By integrating multiple types of markers such as proteomics, inflammation and metabolic dynamics, genomics and epigenetics, exosomes, and metabolites, the present invention comprehensively analyzes from multiple dimensions including protein expression of cells, inflammatory and immune status, genetic changes, intercellular communication, and metabolic activities of cells, comprehensively captures the biological signals of liver metastasis of colorectal cancer, thereby detecting the in vivo microbiological changes during early liver metastasis, improving the sensitivity and specificity of detection, accurately predicting liver metastasis of colorectal cancer, and solving the problem of easy missed diagnosis caused by traditional serum markers.

[0016] 2. By detecting markers such as the ctDNA mutation spectrum, miR-122 and miR-155 in exosomal nucleic acids, the present invention can detect abnormal activities and metastatic tendencies of tumor cells earlier, strive for precious treatment time for patients, and early diagnosis helps to take timely intervention measures to prevent further development and metastasis of tumors, improving the cure rate and survival rate of patients.

[0017] 3. The ctDNA mutation spectrum and miRNA of the present invention reflect the unique molecular characteristics of the tumors of patients, thereby determining the biological behavior of the tumors and the response to treatment. Based on the detection results of the combination of serum markers, a personalized treatment plan is formulated, a more suitable treatment method is selected, unnecessary treatment and adverse drug reactions are avoided, and the treatment effect and the quality of life of patients are improved. Detailed implementation manners

[0018] Next, in combination with the solution of the present invention, the technical solution of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0019] The embodiments of the present invention provide a serum biomarker combination for predicting novel colorectal cancer liver metastasis, including proteomic biomarkers, inflammation and metabolism dynamic biomarkers, genomic and epigenetic biomarkers, exosome biomarkers, metabolite biomarkers. The proteomic biomarkers include Transferrin, Complement C9, Haptoglobin. The inflammation and metabolism dynamic biomarkers include CRP combined with GPS score, lactate dehydrogenase, gamma-glutamyl transpeptidase. The genomic and epigenetic biomarkers include ctDNA mutation profile, miRNA, ctDNA methylation. The exosome biomarker is exosomes secreted by tumor cells, including exosome nucleic acids, exosome proteins. The metabolite biomarkers include polyamines, amino acid metabolites, lipid metabolites; CRP combined with GPS score is used to quantify the systemic inflammation state; the ctDNA mutation profile includes APC, KRAS, BRAF; the miRNA includes miR-21, miR-31; the ctDNA methylation includes SEPT9, RASSF1A; the exosome nucleic acids include miR-122, miR-155; the exosome proteins include CD9, CD63, EpCAM; the polyamines include putrescine, spermine, spermidine; the amino acid metabolites include kynurenine, 5-hydroxytryptamine; the lipid metabolites include phosphatidylcholine, phosphatidylethanolamine.

[0020] Specifically, by detecting the change in the expression level of proteomic biomarkers in serum and analyzing the difference from the expression under normal conditions, the up-regulated expression of Transferrin reflects abnormal iron metabolism in tumor cells. Since tumor cells require a large amount of iron for rapid proliferation, it promotes the formation of the liver metastasis microenvironment and is conducive to the colonization and growth of tumor cells in the liver; the high expression of Complement C9 participates in immune regulation. It inhibits the body's function of clearing tumor cells, enabling tumor cells to evade the surveillance of the body's immune system and thus promoting tumor metastasis; the down-regulation of Haptoglobin indicates a tumor-related chronic inflammatory state. This inflammatory environment provides favorable conditions for tumor metastasis and is related to poor prognosis of liver metastasis. Through the combined detection and comprehensive analysis of these three proteomic biomarkers, the metastatic potential of tumors can be comprehensively reflected from the perspectives of nutrient uptake, immune escape, and inflammatory microenvironment of tumor cells, and the preliminary prediction of the risk of colorectal cancer liver metastasis can be achieved.

[0021] By quantitatively detecting and analyzing dynamic markers of inflammation and metabolism, where the combined CRP and GPS scores are used to quantify the systemic inflammatory state. A high inflammatory level promotes the migration and invasion of tumor cells and increases the risk of liver metastasis; lactate dehydrogenase reflects the metabolic activity of tumors. When tumor cells are metabolically active, an increase in its activity indicates an increased energy demand and altered metabolic pathways of tumor cells, which is closely related to the growth and metastasis of tumors; gamma-glutamyl transpeptidase indicates liver function damage. During colorectal cancer liver metastasis, when the liver is invaded by tumors, it will cause an increase in gamma-glutamyl transpeptidase. By monitoring the changes in these markers, the metabolic state of tumors and the degree of liver function impairment can be understood, the invasiveness and metastatic potential of tumors can be evaluated, and the risk of colorectal cancer liver metastasis can be further predicted and the metastatic invasiveness can be judged.

[0022] By detecting genomic and epigenetic markers, where the detection of ctDNA mutation profiles can identify unique gene mutations in tumor cells. These mutations affect the biological behaviors of tumor cells such as proliferation, differentiation, and metastasis. By detecting mutations in circulating tumor DNA through liquid biopsy, micrometastatic foci can be detected early because even a small number of tumor cells entering the blood circulation can have their mutant ctDNA detected; miRNAs regulate tumor invasion and angiogenesis, and high expression is associated with an increased risk of liver metastasis. By regulating the expression of downstream target genes, it affects the migration and angiogenesis ability of tumor cells; ctDNA methylation changes gene expression and is involved in the process of tumor metastasis. Abnormal methylation can silence or abnormally express related genes, promoting tumor metastasis. By analyzing these markers, the molecular mechanism of tumor metastasis can be revealed at the gene level, providing a basis at the gene level for predicting colorectal cancer liver metastasis, and achieving precise prediction of tumor metastasis risk and individualized treatment guidance.

[0023] By detecting and analyzing exosome markers, where exosomal nucleic acids are involved in tumor cell-to-cell communication. Exosomes secreted by tumor cells carry these nucleic acids into the blood circulation and can be taken up by liver cells, affecting the function of liver cells and creating conditions for tumor cell liver metastasis; exosomal proteins are involved in the interaction between tumor cells and target organ cells. For example, EpCAM can mediate the adhesion of tumor cells to liver cells, promoting the occurrence of liver metastasis. By detecting the content and activity changes of these markers in exosomes, the communication and interaction between tumor cells and surrounding cells can be understood, the trend of tumor cell metastasis to the liver can be judged, and the risk of colorectal cancer liver metastasis can be predicted.

[0024] Through the qualitative and quantitative analysis of metabolite markers, among which polyamines are involved in processes such as cell proliferation, differentiation, and apoptosis, and their levels increase in the serum of patients with liver metastases, indicating active proliferation and metabolism of tumor cells; amino acid metabolites are involved in tumor immune regulation and energy metabolism, and kynurenine can inhibit the function of immune cells, providing conditions for tumor cells to escape immune surveillance; lipid metabolites are involved in the membrane synthesis and signal transduction of tumor cells, and their metabolic changes reflect the growth and metastasis status of tumor cells. By analyzing the changes in these metabolites, the metabolic remodeling and changes in the immune microenvironment during the process of tumor metastasis can be revealed from the perspective of tumor cell metabolism, so as to achieve the prediction of the risk of colorectal cancer liver metastasis.

[0025] When using this serum marker combination, first take 100 - 200 μl of serum samples, and use two-dimensional gel electrophoresis (2-DE) technology for protein separation, and then perform isoelectric focusing to separate on the pH gradient gel strip according to the different isoelectric points of proteins; then perform SDS-PAGE electrophoresis to further separate according to the molecular weight of proteins. Stain the separated protein gel with silver stain or Coomassie brilliant blue, and identify and mark the differentially expressed protein spots through image analysis software.

[0026] Mass spectrometry identification: Cut the differentially expressed protein spots from the gel and put them into a 1.5 ml centrifuge tube, add an appropriate amount of decolorizing solution (containing acetonitrile and ammonium bicarbonate), incubate with shaking at 37 °C for 30 minutes, repeat 2 - 3 times until the gel block becomes colorless, then add an appropriate amount of trypsin solution (10 ng / μl), enzymatically digest at 37 °C for 12 - 16 hours. After the enzymatic digestion is completed, add 5% trifluoroacetic acid solution to terminate the reaction, then extract the enzymatically digested peptide segments with acetonitrile, and then use matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) for identification. Mix the extracted peptide segments with a matrix (such as α-cyano-4-hydroxycinnamic acid), spot them on the target plate, and after the matrix crystallizes, put them into the mass spectrometer for detection. By comparing with protein databases (such as Swiss-Prot, NCBI, etc.), determine the types of differentially expressed proteins, and mainly focus on Transferrin, Complement C9, and Haptoglobin.

[0027] Quantitative detection: An enzyme-linked immunosorbent assay (ELISA) kit was used to quantitatively detect the above three proteins. According to the kit instructions, serum samples and standards (concentration gradients of 0, 10, 50, 100, 500, 1000 ng / ml) were added to a microplate coated with specific antibodies, 100 μl per well, and incubated at 37°C for 1 - 2 hours. After incubation, the liquid in the wells was discarded, and the microplate was washed 3 - 5 times with washing buffer (PBS solution containing Tween-20), with each washing time being 3 - 5 minutes. Then, enzyme-labeled secondary antibody (diluted 1:1000) was added, 100 μl per well, and incubated at 37°C for 1 hour. After washing the microplate again, substrate chromogenic solution (such as TMB substrate) was added, 100 μl per well, and incubated at 37°C in the dark for 15 - 20 minutes. After obvious color development, stop solution (such as 2M sulfuric acid solution) was added to terminate the reaction. The absorbance value was measured at a wavelength of 450 nm using an enzyme-labeled instrument, and the concentration of the protein in the sample was calculated according to the standard curve.

[0028] CRP detection: The concentration of CRP in serum was determined by immunoturbidimetry. Serum samples were mixed with anti-CRP antibodies (polyclonal or monoclonal antibodies) in a certain proportion and incubated at 37°C for 10 - 15 minutes to form antigen-antibody complexes. These complexes cause a turbidity change in the reaction solution, and the increase in turbidity is proportional to the CRP concentration. Then, the turbidity of the reaction solution was detected using an automatic biochemical analyzer, usually at a detection wavelength of 546 nm or 620 nm. The instrument automatically calculates the content of CRP according to the preset standard curve. The standard curve was plotted using a series of CRP standards with different concentrations (such as 0, 5, 10, 20, 50, 100 mg / L), and the same detection method was used for determination. The absorbance value was plotted on the vertical axis and the CRP concentration on the horizontal axis to draw the standard curve.

[0029] Albumin detection: The content of serum albumin was determined by the bromocresol green method using an automatic biochemical analyzer. In a buffer with a pH of 4.2 - 4.5, the lysine residues in albumin molecules bind to bromocresol green to form a blue-green complex, which has a maximum absorption peak at a wavelength of 628 nm. Then, serum samples were mixed with bromocresol green reagent in a certain proportion and incubated at 37°C for 5 - 10 minutes. Then, the absorbance value of the reaction solution at a wavelength of 628 nm was measured on an automatic biochemical analyzer, and the instrument calculates the albumin concentration according to the standard curve. The method for plotting the standard curve is the same as that for CRP detection.

[0030] GPS score calculation: Based on the test results of CRP and albumin, calculate the GPS score according to the following formula: GPS score = 0 (CRP < 10 mg / L and albumin ≥ 35 g / L); GPS score = 1 (CRP ≥ 10 mg / L or albumin < 35 g / L); GPS score = 2 (CRP ≥ 10 mg / L and albumin < 35 g / L).

[0031] LDH and GGT detection: Use an automatic biochemical analyzer and adopt the enzyme kinetics method to detect the activities of LDH and GGT in serum respectively. Taking the LDH detection as an example, in a reaction system containing lactic acid, NAD+, and buffer, LDH catalyzes the dehydrogenation of lactic acid to generate pyruvic acid, and at the same time reduces NAD+ to NADH. NADH has an absorption peak at a wavelength of 340 nm. By monitoring the change rate of the absorbance at 340 nm, the activity of LDH is calculated; the detection principle of GGT is similar. In a reaction system containing γ-glutamyl-p-nitroaniline and glycylglycine, GGT catalyzes the transfer of the γ-glutamyl group to glycylglycine to generate γ-glutamylglycylglycine and p-nitroaniline. p-nitroaniline has an absorption peak at a wavelength of 405 nm. By monitoring the change rate of the absorbance at 405 nm, the activity of GGT is calculated. The instrument automatically calculates the activities of LDH and GGT according to the preset standard curve and reaction kinetic parameters.

[0032] ctDNA extraction: Take 2 - 5 ml of serum samples, add an appropriate amount of proteinase K and lysis buffer, incubate at 56 °C for 30 - 60 minutes to lyse cells and virus particles and release nucleic acids; then use the magnetic bead method to extract ctDNA. Add magnetic bead-capture probes (such as magnetic beads with specific capture oligonucleotides modified on the surface) to the lysate and incubate at room temperature for 15 - 30 minutes. ctDNA will specifically bind to the magnetic beads; then use a magnetic stand to separate the magnetic beads, wash the magnetic beads 3 - 5 times with washing buffer (containing ethanol, Tris-HCl, etc.) to remove impurities, and finally elute ctDNA with elution buffer (such as TE buffer) at 65 °C. Collect the eluate, which is the extracted ctDNA.

[0033] ctDNA Mutation Detection: Next-generation sequencing (NGS) technology is used to sequence ctDNA. The extracted ctDNA is fragmented, and the fragmented ctDNA is subjected to end repair, A-tailing, and ligation of sequencing adapters to construct a sequencing library. The sequencing adapter contains a primer binding site for PCR amplification and a sample-specific tag sequence to distinguish different samples during subsequent sequencing. Then, the library is sequenced using a high-throughput sequencing platform to obtain a large amount of sequencing data. Bioinformatics analysis software is used to align and detect mutations in the sequencing data, and mutation sites of genes such as APC, KRAS, and BRAF are screened out. During the alignment process, the sequencing data is aligned with the human reference genome (such as GRCh38) to determine the position of each sequencing fragment in the genome; during the mutation detection process, statistical analysis and algorithms are used to determine whether there is a mutation at each site, and the mutation type (such as single nucleotide variation, insertion / deletion, etc.) is annotated.

[0034] miRNA Detection: Total RNA, including miRNA, is extracted from serum using a miRNA extraction kit. Stem-loop reverse transcription is used to synthesize cDNA, and then qRT-PCR detection is performed using specific primers and fluorescent quantitative PCR reagents with the cDNA as a template. By detecting the Ct values of miR-21 and miR-31 in different samples, their relative expression levels are calculated according to the 2^-ΔΔCt method.

[0035] ctDNA Methylation Detection: Methylation-specific PCR (MSP) or pyrosequencing technology is used to detect the methylation status of SEPT9 and RASSF1A genes. Taking MSP as an example, first, ctDNA is treated with bisulfite to convert unmethylated cytosine into uracil, while methylated cytosine remains unchanged. Then, specific primers targeting methylated and unmethylated sequences are designed for PCR amplification, and the methylation status of the gene is judged by electrophoresis of the amplification products.

[0036] Exosome Isolation: Exosomes are isolated from serum using the ultracentrifugation method. The serum sample is centrifuged at 3000g for 15 minutes at 4°C to remove cell debris, then centrifuged at 10000g for 30 minutes to remove larger vesicles, and finally ultracentrifuged at 100000g for 70 minutes. The precipitate collected is exosomes.

[0037] Exosome Identification: Nanoparticle tracking analysis (NTA) technology is used to determine the size distribution and concentration of exosomes. The size of exosomes is usually between 30 - 150 nm. Transmission electron microscopy (TEM) is used to observe the morphology of exosomes, and exosomes show a typical cup-shaped or spherical structure.

[0038] Exosomal nucleic acid detection: Use an RNA extraction kit to extract RNA from exosomes, and detect the expression levels of miR-122, miR-155, etc. by qRT-PCR. The method is the same as that for serum miRNA detection.

[0039] Exosomal protein detection: Detect the expression of exosomal surface proteins CD9, CD63, and EpCAM by Western blot or flow cytometry. Taking Western blot as an example, lyse the exosomes to extract proteins, perform SDS-PAGE electrophoresis separation, then transfer the proteins to a PVDF membrane, and perform immunoblotting with specific antibodies. Detect the expression of the target protein by chemiluminescence.

[0040] Sample pretreatment: Take 100 - 200 μl of serum sample, add an appropriate amount of methanol or acetonitrile for protein precipitation. After vortex oscillation, centrifuge at 12000 g for 15 minutes at 4°C, and take the supernatant. The supernatant is used for subsequent analysis after concentration, derivatization, etc.

[0041] Detection and analysis: Use liquid chromatography-mass spectrometry (LC-MS) or gas chromatography-mass spectrometry (GC-MS) to qualitatively and quantitatively analyze polyamines (putrescine, spermine, spermidine), amino acid metabolites (kynurenine, 5-hydroxytryptamine), and lipid metabolites (phosphatidylcholine, phosphatidylethanolamine). Determine the types and contents of metabolites by comparing with standards and searching the database.

[0042] Standardize the detection data of all markers to eliminate the differences caused by different detection methods and units. Common standardization methods include Z-score standardization, normalization, etc.; use dimensionality reduction methods such as principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) to preprocess the data, screen out the markers most strongly correlated with colorectal cancer liver metastasis, reduce the data dimension, and improve the accuracy and stability of the model; then use machine learning algorithms such as logistic regression (LR), support vector machine (SVM), and random forest (RF) to construct a colorectal cancer liver metastasis prediction model. Divide the dataset into a training set and a test set, usually in a ratio of 7:3 or 8:2. Use the training set data to train the model, adjust the model parameters, and then use the test set data to verify and evaluate the model. Evaluate the performance of the model by calculating indicators such as accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve (AUC). Use methods such as five-fold cross-validation or leave-one-out method to verify the model multiple times to ensure the reliability and generalization ability of the model.

[0043] According to the constructed prediction model, the serum samples of new colorectal cancer patients are detected and analyzed to predict the risk of liver metastasis in patients. The risks are divided into three levels: high, medium, and low, providing a reference for clinicians to formulate treatment plans. For patients predicted to be at high risk, it is recommended to adopt more aggressive treatment strategies, such as preoperative neoadjuvant chemotherapy, targeted therapy, etc., to reduce the risk of liver metastasis. For patients predicted to be at low risk, the treatment intensity can be appropriately reduced to avoid over-treatment, and then by monitoring the dynamic changes of serum markers during the treatment of patients, the treatment effect and prognosis are evaluated. If the marker level decreases, it indicates that the treatment is effective; if the marker level increases, it may indicate tumor recurrence or metastasis, and the treatment plan needs to be adjusted in a timely manner.

[0044] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A novel serum marker combination for predicting liver metastasis of colorectal cancer, including proteomic markers, inflammatory and metabolic dynamic markers, genomic and epigenetic markers, exosome markers, and metabolite markers, characterized in that: The proteomic markers include Transferrin, Complement C9, and Haptoglobin; the inflammatory and metabolic dynamic markers include CRP combined with GPS score, lactate dehydrogenase, and glutamate transpeptidase; the genomic and epigenetic markers include ctDNA mutation spectrum, miRNA, and ctDNA methylation; the exosome markers are exosomes secreted by tumor cells, including exosome nucleic acids and exosome proteins; and the metabolite markers include polyamines, amino acid metabolites, and lipid metabolites.

2. A novel serum marker combination for predicting liver metastasis of colorectal cancer according to claim 1, characterized in that: The CRP combined with the GPS score is used to quantify the systemic inflammatory status.

3. A novel serum marker combination for predicting liver metastasis of colorectal cancer according to claim 1, characterized in that: The ctDNA mutation spectrum includes APC, KRAS, and BRAF.

4. A novel serum marker combination for predicting liver metastasis of colorectal cancer according to claim 1, characterized in that: The miRNAs include miR-21 and miR-31.

5. The novel serum marker combination for predicting liver metastasis of colorectal cancer according to claim 1, characterized in that: The ctDNA methylation includes SEPT9 and RASSF1A.

6. A novel serum marker combination for predicting liver metastasis of colorectal cancer according to claim 1, characterized in that: The exosomal nucleic acids include miR-122 and miR-155.

7. A novel serum marker combination for predicting liver metastasis of colorectal cancer according to claim 1, characterized in that: The exosomal proteins include CD9, CD63, and EpCAM.

8. The novel serum marker combination for predicting liver metastasis of colorectal cancer according to claim 1, characterized in that: The polyamine substances include putrescine, spermine and spermidine.

9. A novel serum marker combination for predicting liver metastasis of colorectal cancer according to claim 1, characterized in that: The amino acid metabolites include kynurenine and 5-hydroxytryptamine.

10. The novel serum marker combination for predicting liver metastasis of colorectal cancer according to claim 1, characterized in that: The lipid metabolites include phosphatidylcholine and phosphatidylethanolamine.